HP XP P9500 User Manual

Hitachi dynamic link manager software user guide for solaris (6.6) (hit5205-96011, november 2011).
Hide thumbs
   
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574
Hitachi Command Suite
Dynamic Link Manager Software
User Guide (for Solaris(R))
Document Organization
Product Version
Contents
MK-92DLM114-21

Advertising

   Also See for HP XP P9500

   Related Manuals for HP XP P9500

   Summary of Contents for HP XP P9500

  • Page 1

    Hitachi Command Suite Dynamic Link Manager Software User Guide (for Solaris(R)) Document Organization Product Version Getting Help Contents MK-92DLM114-21...

  • Page 2

    © 2011 Hitachi, Ltd., Hitachi Data Systems Corporation. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or stored in a database or retrieval system for any purpose without the express written permission of Hitachi, Ltd.

  • Page 3: Table Of Contents

    Contents Preface....................xiii Intended audience....................xiv Product version.......................xiv Release notes......................xiv Document revision level...................xiv Document organization....................xiv Referenced documents.....................xv Document conventions.....................xv Convention for storage capacity values..............xvii Getting help......................xviii Comments......................xviii 1 Overview of HDLM................1-1 What is HDLM?.......................1-2 HDLM Features.......................1-3 2 HDLM Functions...................2-1 Devices Managed by HDLM..................2-3 System Configuration....................2-3 LU Configuration.....................2-5 Program Configuration....................2-6...

  • Page 4: Table Of Contents

    Path Status Transition..................2-19 The Online Path Status................2-19 The Offline Path Status................2-19 Status Transitions of a Path..............2-20 Intermittent Error Monitoring (Functionality When Automatic Failback Is Used)..2-21 Checking Intermittent Errors................2-22 Setting Up Intermittent Error Monitoring............2-22 Intermittent Error Monitoring Actions...............2-22 When an Intermittent Error Occurs............2-22 When an Intermittent Error Does Not Occur..........2-23 When the Conditions for an Intermittent Error Are Changed During Error Monitoring....................2-24...

  • Page 5: Table Of Contents

    Notes on Command Execution.................3-41 Notes on the Disk Label...................3-41 Installing HDLM....................3-41 Preparations for a New Installation of HDLM.............3-41 Performing Operations on Devices to Be Managed by HDLM.....3-41 Apply Solaris Patches................3-43 Set Up the Hardware................3-43 Set Up the /kernel/drv/sd.conf File............3-44 Switch the Kernel Mode ................3-44 Set Up the /etc/system File..............3-45 Set Up the /etc/syslog.conf File..............3-45 Set Up VxVM..................3-46...

  • Page 6: Table Of Contents

    Excluding the Prepared LUs from HDLM Management......3-113 Configuring an Environment and then Uninstalling HDLM......3-113 Checking the Path Configuration................3-119 Setting Up HDLM Functions..................3-120 Checking the Current Settings................3-120 Setting Up the HDLM Functions..............3-120 Setting Up Load Balancing..............3-121 Setting Up Path Health Checking............3-122 Setting Up the Automatic Failback Function...........3-122 Setting Up Intermittent Error Monitoring..........3-123 Setting the Error Log Collection Level............3-123...

  • Page 7: Table Of Contents

    Overview of HDLM Uninstallation..............3-164 Preparations for HDLM Uninstallation..............3-165 Performing Operations on HDLM-Managed Devices........3-165 Remove Sun Cluster Settings..............3-166 Remove VCS Settings................3-169 Remove VxVM Settings................3-169 Remove SDS Settings................3-169 Remove SVM Settings................3-170 Uninstalling HDLM..................3-171 Uninstalling HDLM from the Local Boot Disk Environment.......3-171 Uninstalling HDLM from the Boot Disk Environment........3-173 Uninstalling HDLM from an LDoms Environment........3-173 Settings Needed After HDLM Uninstallation.............3-178 VxVM Settings..................3-178...

  • Page 8: Table Of Contents

    Viewing HDLM Component Information............4-14 Starting and Stopping the HDLM Manager...............4-15 Starting the HDLM Manager................4-15 Stopping the HDLM Manager................4-15 HDLM Resident Processes..................4-16 Changing the Configuration of the HDLM Operating Environment......4-16 Precautions Regarding Changes to the Configuration of an HDLM Operating Environment....................4-17 Changing the Configuration of a System that Uses HDLM......4-17 When the Path Configuration Is Changed..........4-17 Switching the Kernel Mode..............4-19...

  • Page 9: Table Of Contents

    What To Do for a Program Error................5-7 Examining the Messages................... 5-8 Obtaining Program Information................. 5-8 What To Do for the Program Error..............5-8 Contacting Your HDLM Vendor or Maintenance Company........5-9 What To Do for Other Errors..................5-9 6 Command Reference................6-1 Overview of the HDLM Command dlnkmgr..............6-2 clear (Returns the Path Statistics to the Initial Value)..........

  • Page 10: Table Of Contents

    To display path information..............6-38 To display LU information..............6-50 To display HBA port information.............6-58 To display CHA port information.............6-60 To display corresponding information about an HDLM device, sd or ssd device, and LDEV....................6-61 To display view operation format............6-62 monitor (Displays I/O Information at a Specified Interval)........6-63 Format......................6-63 To display I/O information for each HBA port..........6-63 To display I/O information for each CHA port..........6-63...

  • Page 11: Table Of Contents

    8 Messages.....................8-1 Before Viewing the List of Messages.................8-3 Format and Meaning of Message IDs..............8-3 Terms Used in Messages and Message Explanations..........8-3 Components That Output Messages to Syslog.............8-3 KAPL01001 to KAPL02000..................8-4 KAPL03001 to KAPL04000..................8-25 KAPL04001 to KAPL05000..................8-28 KAPL05001 to KAPL06000..................8-35 KAPL06001 to KAPL07000..................8-42 KAPL07001 to KAPL08000..................8-45 KAPL08001 to KAPL09000..................8-46...

  • Page 12

    Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 13

    Preface This document describes how to use the Hitachi Dynamic Link Manager. □ Intended audience □ Product version □ Release notes □ Document revision level □ Document organization □ Referenced documents □ Document conventions □ Convention for storage capacity values □...

  • Page 14: Overview Of Hdlm

    Intended audience This manual is intended for system administrators who use Hitachi Dynamic Link Manager (HDLM) to operate and manage storage systems. The readers of this manual must have a basic knowledge of the following areas: • Solaris and its management functionality •...

  • Page 15: A Sun Cluster 3.2 Commands, Sun Cluster 3.2 Commands

    Chapter Description stop the HDLM manager. This chapter also describes how to configure an environment to properly operate HDLM, such as changing the HDLM management-target devices that connect paths or replacing the hardware that makes up a path. describes how to check path information by using the Windows management tool.

  • Page 16

    • General font conventions • Conventions in syntax explanations These conventions are described below. General font conventions The following table lists the general font conventions: Font Convention Bold Bold type indicates text on a window, other than the window title. Such text includes menus, menu options, buttons, radio box options, or explanatory labels.

  • Page 17

    Example font or Convention symbol database-name This font style marks a placeholder that indicates where appropriate characters are to be entered in an actual command. Bold code-font characters indicate the abbreviation for a command. Underlined characters indicate the default value. perm Square brackets enclose an item or set of items whose specification is optional.

  • Page 18: Getting Help

    Logical capacity unit Value 1 MB 1,024 KB or 1,024 bytes 1 GB 1,024 MB or 1,024 bytes 1 TB 1,024 GB or 1,024 bytes 1 PB 1,024 TB or 1,024 bytes 1 EB 1,024 PB or 1,024 bytes Getting help The Hitachi Data Systems customer support staff is available 24 hours a day, seven days a week.

  • Page 19: Overview Of Hdlm

    Overview of HDLM HDLM is a software package that manages paths between a host and a storage system. HDLM is designed to distribute loads across multiple paths and will switch a given load to another path if there is a failure in the path that is currently being used, thus improving system reliability.

  • Page 20: What Is Hdlm

    What is HDLM? With the widespread use of data warehousing and increasing use of multimedia data, the need for high-speed processing of large volumes of data on networks has rapidly grown. To satisfy this need, networks dedicated to the transfer of data, such as SANs, are now being used to provide access to storage systems.

  • Page 21: Hdlm Features

    HDLM Features HDLM features include the following: The ability to distribute a load across multiple paths. This is also known as load balancing. When a host is connected to a storage system via multiple paths, HDLM can distribute the load across all the paths. This prevents one, loaded down path from affecting the processing speed of the entire system.

  • Page 22

    Overview of HDLM Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 23: Hdlm Functions

    HDLM Functions This chapter describes the various functions that are built into HDLM. Before the function specifications are explained though, this chapter will go into detail about the HDLM management targets, system configuration, and basic terms that are necessary to know to effectively operate HDLM. After that, the rest of the chapter focus on describing all the HDLM functions, including the main ones: load distribution across paths and path switching.

  • Page 24

    □ Integrated HDLM management using Global Link Manager □ Cluster Support HDLM Functions Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 25: Devices Managed By Hdlm, System Configuration

    Devices Managed by HDLM Below is a list of devices that can or cannot be managed by HDLM. The devices that can be managed by HDLM are called HDLM management-target devices. HDLM management-target devices: The following devices are from the storage systems listed in Section What is HDLM? on page 1-2:...

  • Page 26

    paths. The ID that HDLM assigns to each path (physical path) is called a AutoPATH_ID. A path is also sometimes called a managed object. Figure 2-1 HDLM System Configuration on page 2-4 shows the HDLM system configuration. Figure 2-1 HDLM System Configuration Table 2-1 HDLM System Components on page 2-4 lists the HDLM system components.

  • Page 27

    LU Configuration After you have properly installed HDLM, the LU configuration will change as follows: Before the installation of HDLM: The host recognizes that an sd or ssd device is connected to each physical path. Thus, a single LU in the storage system is recognized as the same number of LUs as that of physical paths.

  • Page 28: Program Configuration

    Figure 2-2 LU Configuration Recognized by the Host After the Installation of HDLM Table 2-2 Components Recognized by the Host on page 2-6 lists the components recognized by the host. Table 2-2 Components Recognized by the Host Components Description An LU that the host recognizes via the HDLM driver. It is called a host LU.

  • Page 29

    Figure 2-3 Configuration of the HDLM Programs Table 2-3 Function of HDLM Programs on page 2-7 lists and describes the functions of these programs. Table 2-3 Function of HDLM Programs Program name Functions HDLM command Provides the dlnkmgr command, which enables you to: •...

  • Page 30: Position Of The Hdlm Driver And Hdlm Device

    Program name Functions • Configure the HDLM operating environment • Request path health checks and automatic failbacks to be performed • Collect error log data HDLM alert driver Reports the log information collected by the HDLM driver to the HDLM manager. The driver name is dlmadrv. HDLM driver Controls all the HDLM functions, manages paths, and detects errors.

  • Page 31: Logical Device Files For Hdlm Devices

    Figure 2-4 Position of the HDLM Driver and HDLM Devices Logical Device Files for HDLM Devices When you install HDLM, a logical device file to be used by HDLM will be created for each LU on a per-Dev (slice) basis. Setting this logical device file name in an application, such as volume management software, enables the application to access an LU by using the HDLM function.

  • Page 32: Distributing A Load Using Load Balancing

    and for one of the Dev (slices) in that LU, the corresponding logical device file names of the sd or ssd devices are c2t1d1s0 and c3t2d1s0. In this case, when you install HDLM, these logical device files will be deleted. Then, a logical device file that has a different controller number, such as c4t1d1s0, is created for the HDLM device.

  • Page 33

    Figure 2-5 Flow of I/O Data When the Load Balancing Function Is Not Used When the load balancing function is not used, I/O operations converge onto a single path (A). The load on that one physical path (A) will cause a bottleneck, which might cause problems with system performance.

  • Page 34: Paths To Which Load Balancing Is Applied

    Figure 2-6 Flow of I/O Data When the Load Balancing Function Is Used When the load balancing function is used, I/O operations are distributed via multiple physical paths (A, B, C, and D). This helps to prevent problems with system performance and helps prevent bottlenecks from occurring. Paths to Which Load Balancing Is Applied This section describes, for each type of storage system, the paths to which the load balancing function is applied.

  • Page 35: Series, Or Hitachi Sms, Load Balancing Algorithms

    remaining usable owner paths. When all owner paths cannot be used, load balancing is performed among the non-owner paths. For the example in Figure 2-7 Overview of Load Balancing on page 2-13, suppose that in the owner controller of LU0 is CHA0. When the LU is accessed, the load is balanced between the two physical paths A and B, which are both owner paths.

  • Page 36

    • The Least I/Os algorithm • The Extended Least I/Os algorithm • The Least Blocks algorithm • The Extended Least Blocks algorithm The above algorithms are divided into two categories, which differ in their processing method. The following describes both of these processing methods: The Round Robin, Least I/Os, and Least Blocks algorithms These algorithms select which path to use every time an I/O is issued.

  • Page 37: Performing Failovers And Failbacks Using Path Switching

    Table 2-4 Features of the Load Balancing Algorithms Algorithm type Algorithm features • These types of algorithms are most effective when a lot of Round Robin discontinuous, non-sequential I/Os are issued. • Least I/Os • Least Blocks • Extended Round If the I/O data is from something like a read request and is Robin generally sequential with the previous I/Os, an improvement...

  • Page 38: Automatic Path Switching, Automatic Failovers

    • An error occurred on the path. • A user executed the HDLM command's offline operation. For details on the offline operation, see offline (Places Paths Offline) on page 6-6. For details on path statuses and the transitions of those statuses, see Path Status Transition on page 2-19.

  • Page 39

    that path is placed offline, the switching destination can come from any of the other three physical paths (B, C, or D). Paths are switched in units of physical paths. Therefore, if an error occurs in a path, HDLM switches all the other paths that run through the same physical path.

  • Page 40: Manual Path Switching

    Series, or Hitachi SMS is being used, all the paths are owner paths. Therefore, if an owner path recovers from an error and HDLM automatically places the recovered path online, the path in use will not be switched to the recovered owner path.

  • Page 41: Path Status Transition, The Online Path Status, The Offline Path Status

    the path status is changed to online, HDLM selects the path to use in the same way as for automatic path switching. When the Thunder 9500V Series, or Hitachi AMS/WMS series is being used, HDLM selects the path to use from online owner paths, and then from online non-owner paths.

  • Page 42: Status Transitions Of A Path

    Status Transitions of a Path Figure 2-9 Path Status Transitions on page 2-20 shows the status transitions of a path. Figure 2-9 Path Status Transitions Legend: Online operation: Online operation performed by executing the dlnkmgr command's online operation. Offline operation: Offline operation performed by executing the dlnkmgr command's offline operation.

  • Page 43: Intermittent Error Monitoring (functionality When Automatic Failback Is Used)

    When an Online or Offline(E) path exists among the paths that access the same LU. One of the Offline(E) paths is changed to the Online(E) path. When an Offline(E) path exists among the paths that access the same LU. The last available online path for each LU cannot be placed offline by executing the offline operation.

  • Page 44: Checking Intermittent Errors, Setting Up Intermittent Error Monitoring, Intermittent Error Monitoring Actions

    Checking Intermittent Errors You can check the paths in which intermittent errors have occurred by viewing the execution results of the HDLM command's view operation. For details on the view operation, see view (Displays Information) on page 6-30. Setting Up Intermittent Error Monitoring When you use the intermittent error function, you can enable or disable the function.

  • Page 45: When An Intermittent Error Does Not Occur

    Figure 2-10 Action What Will Happen When an Intermittent Error Occurs on a Path on page 2-23 shows what will happen when an intermittent error occurs. For this example, the path is determined to have an intermittent error when the error occurs 3 or more times within 30 minutes. The events that occur are described by using the time arrows.

  • Page 46

    Figure 2-11 What Will Happen When an Intermittent Error Does Not Occur on a Path As shown in Figure 2-11 What Will Happen When an Intermittent Error Does Not Occur on a Path on page 2-24, normally, the count for the number of times that an error occurs is started after the path is first recovered from an error by using the automatic failback function.

  • Page 47: When A User Changes The Intermittent Error Information

    Figure 2-12 What Will Happen When Conditions Are Changed During Error Monitoring When a User Changes the Intermittent Error Information The following might be reset when a user changes any of the values set for the intermittent error or the path status: the number of errors that have already been counted during error monitoring, the amount of time that has passed since error monitoring has started, and the information about whether an intermittent error has occurred.

  • Page 48

    Number of Information errors and time about paths passed since User operation not subject to error automatic monitoring failback started changing the conditions) while intermittent error monitoring is being performed Changing the conditions for (Not applicable) Inherited an intermittent error outside (Not counted.) the intermittent error monitoring...

  • Page 49: Detecting Errors By Using Path Health Checking, Error Management

    Detecting Errors by Using Path Health Checking HDLM can check the status of paths to which I/Os are not issued, at regular intervals, and detect errors. This function is called path health checking. Without path health checking, an error is not detected unless I/O is issued because the system only checks the path status when I/O is issued.

  • Page 50: Types Of Collected Logs

    Figure 2-13 Flow of Data When Collecting Error Information Logs might be collected in layers below HDLM, such as for the SCSI driver. For more details, see the Solaris documentation. Types of Collected Logs HDLM collects information on detected errors and trace information in the integrated trace file, trace file, error logs, log for the dlmcfgmgr utility for managing the HDLM configuration, and syslog.

  • Page 51: Error Information

    Table 2-6 Types of Error Information on page 2-29 describes the types of error information. Table 2-6 Types of Error Information Log name Description Output destination Integrated trace file Operation logs for the HDLM The default file path is / command is collected.

  • Page 52: Filtering Of Error Information

    When you want to configure the system so that HDLM messages are output to syslog, specify user for the facility in the /etc/syslog.conf file. The following shows an example where the system function name is user, and messages at the info level or higher are output to the /tmp/ syslog.user.log file: user.info /tmp/syslog.user.log...

  • Page 53: Collecting Audit Log Data

    Collects error information from the Warning level and higher. ¢ Collects error information from the Information level and higher. ¢ Collects error information from the Information level and higher ¢ (including maintenance information). Collection levels for log information in trace files: Outputs no trace information ¢...

  • Page 54

    Category Explanation • Startup and termination of software running on storage systems, software running on SVPs (service processors), and Hitachi Command Suite products An abnormal hardware or software event, including: Failure • Hardware errors • Software errors (such as memory errors) An event indicating the linkage status between devices: LinkStatus •...

  • Page 55: Categories And Audit Events That Hdlm Can Output To The Audit Log

    Category Explanation An event indicating an abnormal state such as exceeding a AnomalyEvent threshold, including: • Exceeding a network traffic threshold • Exceeding a CPU load threshold • Reporting that the temporary audit log data saved internally is close to its maximum size limit or that the audit log files have wrapped back around to the beginning An event indicating an occurrence of abnormal...

  • Page 56

    Severity Category Explanation Audit event Message ID The I/O KAPL15114-I information monitoring function stopped. The I/O KAPL15115- information monitoring function terminated. Startup of the KAPL15060-I DLMgetras utility Termination of the KAPL15061-I DLMgetras utility Administrator Permission has not KAPL15111- Authentication or end-user been granted to authentication execute the HDLM...

  • Page 57

    Severity Category Explanation Audit event Message ID information was successful. An attempt to KAPL15108-E display program information failed. An attempt to KAPL15109-I display HDLM management- target information was successful. An attempt to KAPL15110-E display HDLM management- target information failed. Processing of the KAPL15020-I dlmcfgmgr -a command was...

  • Page 58: Requirements For Outputting Audit Log Data, Destination And Filtering Of Audit Log Data

    The severity levels are as follows: 3: Error, 4: Warning, 6: Informational If you use Ctrl + C to cancel the DLMgetras utility for collecting HDLM error information, audit log data indicating that the DLMgetras utility has terminated will not be output. Requirements for Outputting Audit Log Data HDLM can output audit log data when all of the following conditions are satisfied:...

  • Page 59: Audit Log Data Formats

    • Use the HDLM command's set operation to specify local0 for the audit log facility: You can also filter the audit log output by specifying a severity level and type for the HDLM command's set operation. Filtering by severity: The following table lists the severity levels that can be specified. Table 2-10 Severity Levels That Can Be Specified Correspondence with syslog Severity...

  • Page 60

    The format of message-section: common-identifier,common-specification-revision-number,serial- number,message-ID,date-and-time,entity-affected,location- affected,audit-event-type,audit-event-result,subject-ID-for-audit-event- result,hardware-identification-information,location-information,location- identification-information,FQDN,redundancy-identification- information,agent-information,host-sending-request,port-number- sending-request,host-receiving-request,port-number-receiving- request,common-operation-ID,log-type-information,application- identification-information,reserved-area,message-text Up to 950 bytes of text can be displayed for each message-section. Table 2-11 Items Output in the Message Section Explanation Item Common identifier Fixed to CELFSS Common specification Fixed to 1.1 revision number...

  • Page 61: Integrated Hdlm Management Using Global Link Manager

    Explanation Item Host receiving request Name of the host receiving a request Port number receiving Number of the port receiving a request request Common operation ID Operation serial number in the program Log type information Fixed to BasicLog Application Program identification information identification information Reserved area...

  • Page 62: Cluster Support

    Figure 2-14 Example System Configuration Using HDLM and Global Link Manager Cluster Support HDLM can also be used in cluster configurations. For details on cluster software supported by HDLM, the supported Solaris version, and usable volume management software, see Combinations of Cluster Software and Volume Managers Supported by HDLM on page 3-7.

  • Page 63

    displayed by executing the following command, are the same, the versions of HDLM are the same: # /opt/DynamicLinkManager/bin/dlnkmgr view –sys 2-41 HDLM Functions Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 64

    2-42 HDLM Functions Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 65: Creating An Hdlm Environment

    Creating an HDLM Environment This chapter describes the procedures for setting up an HDLM environment and the procedure for canceling those settings. Make sure that HDLM is installed and the functions have been set up. Set up volume groups and cluster software to suit your operating environment. □...

  • Page 66

    □ Creating File Systems for HDLM (When Volume Management Software Is Not Used) □ Setting Up VxVM □ Setting Up SDS □ Setting Up SVM □ Setting Up VCS □ Uninstalling HDLM Creating an HDLM Environment Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 67: Hdlm System Requirements, Hosts And Oss Supported By Hdlm, Storage Systems Supported By Hdlm

    HDLM System Requirements Check the following before installing HDLM: Hosts and OSs Supported by HDLM You can install HDLM on Sun SPARC series computer on which an OS listed in Table 3-1 Applicable OSs for the host on page 3-3 is running.

  • Page 68: Hbas

    Hitachi Universal Storage Platform 100 ¢ Hitachi Universal Storage Platform 600 ¢ Hitachi Universal Storage Platform 1100 ¢ Hitachi Universal Storage Platform V ¢ Hitachi Universal Storage Platform VM ¢ Hitachi Virtual Storage Platform ¢ HP StorageWorks P9500 Disk Array ¢...

  • Page 69: When Handling Intermediate Volumes Managed By Hitachi Rapidxchange, Cluster Software Supported By Hdlm

    When Handling Intermediate Volumes Managed by Hitachi RapidXchange When you exchange data by using intermediate volumes managed by Hitachi RapidXchange, the following version of File Access Library and File Conversion Utility (FAL/FCU) is required: • For Lightning 9900V Series 01-03-56/20 or later •...

  • Page 70: Volume Manager Supported By Hdlm

    Platform V/VM, or Virtual Storage Platform storage systems are connected in a Solaris 10 environment. Note that only the failover and parallel service groups are supported. The hybrid service group is not supported. Can be used when the prerequisite Sun Cluster patches are applied. The following configurations are not supported: A configuration in which Oracle RAC uses the LU that specified the EFI ¢...

  • Page 71: Combinations Of Cluster Software And Volume Managers Supported By Hdlm

    Combinations of Cluster Software and Volume Managers Supported by HDLM For the Sun Cluster or VCS Environment Table 3-3 Combinations of related programs supported by HDLM on page lists the combinations of cluster software and volume managers that are supported by HDLM. Table 3-3 Combinations of related programs supported by HDLM Cluster Volume manager...

  • Page 72

    Cluster Volume manager #2 #3 #5 None VCS 5.0 VxVM 5.0 Solaris 10 None None #4 #8 #9 SVM 1.0 #5 #9 VxVM 4.1 #5 #9 VxVM 5.0 None #1 #5 #9 Sun Cluster 3.1 (8/05) #6 #10 SVM 1.0 VxVM 4.1 #1 #11 Sun Cluster 3.2...

  • Page 73

    When a failure occurs on one of the nodes in a two-node configuration ¢ running Sun Cluster and the LU cannot be accessed When the SDS 4.2.1 or SVM 1.0 shared diskset is being used in an ¢ environment without Sun Cluster Does not support the I/O fencing function.

  • Page 74: When Creating An Oracle9i Rac Environment

    For a two-node configuration, the pathcount setting is only supported for the SCSI protocol (fencing protocol) of the storage device. For details on how to specify SCSI protocols for storage devices, refer to the Sun Cluster manual. For the EFI label or ZFS, only two-node configurations are supported. Only two-node configurations are supported.

  • Page 75: When Creating An Oracle Rac 10g Environment

    Program Remarks • SUNWscucm • SUNWudlm • SUNWudlmr Oracle9i 9.2.0.8.0 RAC is bundled with Oracle9i. Oracle UNIX Distributed Lock Required packages: Manager 3.3.4.8 ORCLudlm Volume Manager None (Specify an HDLM raw device by the device ID of Sun Cluster) When Creating an Oracle RAC 10g Environment Required programs Table 3-5 Programs required to create an Oracle RAC 10g environment (For Solaris 8 or Solaris 9) on page 3-11...

  • Page 76

    Configurati Program Remarks the documentation for Oracle RAC 10g. Configuration Solaris 8 or Solaris 9 • In Solaris 8, use Update 7 or later. • In Solaris 9, use Update 6 or later. Oracle RAC 10g Oracle 10g Database 10.2.0.1.0 Cluster Oracle Clusterware 10.2.0.1.0...

  • Page 77

    Configurati Program Remarks Volume None (Specify an HDLM Manager raw device) Configuration Solaris 9 Oracle RAC 10g Oracle 10g Database Only configurations that 10.2.0.2.0 consist of three or more nodes are supported. Cluster Sun Cluster 3.1 8/05 and Oracle Clusterware 10.2.0.2.0 Volume None (Specify an HDLM...

  • Page 78

    Configurati Program Remarks • ASM is used as the disk memory area for files and recovery files of the Oracle database. In Oracle RAC 10g, HDLM devices can be used following the same procedures as for disk devices. For details on how to install ASM, refer to the documentation for Oracle RAC 10g.

  • Page 79

    Configurati Program Remarks • ASM is used as the disk memory area for files and recovery files of the Oracle database. In Oracle RAC 10g, HDLM devices can be used following the same procedures as for disk devices. For details on how to install ASM, refer to the documentation for Oracle RAC 10g.

  • Page 80

    Configurati Program Remarks Configuratio Solaris 10 Oracle RAC Oracle 10g Database 10.2.0.3.0 Cluster Oracle Clusterware 10.2.0.3.0 Volume None (Specify an HDLM raw Manager device) Configuratio Solaris 10 Oracle RAC Oracle 10g Database 10.2.0.3.0 Cluster Oracle Clusterware 10.2.0.3.0 Volume • ASM is bundled with Manager Oracle RAC 10g.

  • Page 81

    Configurati Program Remarks and recovery files. For the disk device used by ASM, specify the Sun Cluster device ID. For details on how to use ASM, refer to the documentation for Oracle RAC 10g. Configuratio Solaris 10 n 12 Oracle RAC Oracle 10g Database Only two-node 10.2.0.3.0...

  • Page 82

    Configurati Program Remarks Configuratio Solaris 10 n 15 Oracle RAC Oracle 10g Database 10.2.0.3.0 Cluster Oracle Solaris Cluster 3.3 and Oracle Clusterware 10.2.0.3.0 Volume None (Specify an HDLM raw Manager device from the Sun Cluster device ID) Configuratio Solaris 10 n 16 Oracle RAC Oracle 10g Database...

  • Page 83

    Configurati Program Remarks Oracle RAC Oracle 10g Database Only two-node 10.2.0.4.0 configurations are supported. Cluster Oracle Solaris Cluster 3.3 and Oracle Clusterware 10.2.0.4.0 Volume VxVM 5.1 cluster Allocates memory areas, Manager shared among nodes, such functionality as Oracle database files, SPFILE, REDO log files, Oracle Cluster Registry, and voting disks, to the VxVM...

  • Page 84

    Required patches Table 3-7 Patches required to create an Oracle RAC 10g environment (For Solaris 8) on page 3-20,Table 3-8 Patches required to create an Oracle RAC 10g environment (For Solaris 9) on page 3-21 show patches that are provided by Sun Microsystems Inc. and are required to create an Oracle RAC 10g environment.

  • Page 85

    Table 3-8 Patches required to create an Oracle RAC 10g environment (For Solaris 9) Target program Patch ID Timing for applying Oracle RAC 10g 112233-11 or later Apply the patch before installing Oracle RAC 10g. 111722-04 or later 113801-12 or later It is necessary only for a configuration where Sun Cluster 3.1 8/05 is used as the cluster.

  • Page 86

    When using Oracle RAC 10g version 10.2.0.2.0 or later (when Sun ¢ Cluster or Storage Foundation for Oracle RAC is used): Change the value of MISSCOUNT to match the type of storage system. To do so, use the following table to obtain the value to be specified, and then change the current value to a value equal to or greater than the value you have obtained.

  • Page 87

    For details on how to change MISSCOUNT and DISKTIMEOUT, contact the company with which you have a contract for Oracle Support Services. Note that when you uninstall HDLM from the above configuration, you must reset the values of MISSCOUNT and DISKTIMEOUT to their original values.

  • Page 88: When Creating An Oracle Rac 11g Environment

    For the created alias device file, set the owner, group, and access permission mode by using the following command. The owner, group, and access permission mode to be set differs depending on the usage purpose of Oracle RAC 10g corresponding to that device. For details on this point, refer to the Oracle documentation.

  • Page 89

    Configurati Program Remarks Volume • ASM is bundled with Manager Oracle RAC 11g. • ASM is used as the disk memory area for the Oracle database files and recovery files. In Oracle RAC 11g, HDLM devices can be used following the same procedures as for disk devices.

  • Page 90

    Configurati Program Remarks Volume None (Specify an HDLM Manager raw device) Configuratio Solaris 10 Oracle RAC 11g Oracle 11g Database Only two-node configurations 11.1.0.6.0 are supported. Cluster Sun Cluster 3.2 and Oracle Clusterware 11.1.0.6.0 Volume • ASM is bundled with Manager Oracle RAC 11g.

  • Page 91

    Configurati Program Remarks documentation for Oracle RAC 11g. Configuratio Solaris 10 Oracle RAC 11g Oracle 11g Database 11.1.0.7.0 Cluster Oracle Clusterware 11.1.0.7.0 Volume None (Specify an HDLM Manager raw device) Configuratio Solaris 10 n 10 Oracle RAC 11g Oracle 11g Database 11.1.0.7.0 Cluster Oracle Clusterware...

  • Page 92

    You must apply MP3 or later. Note • When a host and an Oracle RAC 11g voting disk are connected by multiple paths, HDLM performs failover processing for those paths (in the same way as for normal paths) when an I/O timeout occurs for one of the paths.

  • Page 93

    Table 3-14 Formula for Calculating DISKTIMEOUT Number of paths connecte Formula for obtaining the value of Storage system type d to the DISKTIMEOUT voting disk • Lightning 9900 series 3 or less You do not need to change the value of DISKTIMEOUT.

  • Page 94: Virtualization Environments Supported By Hdlm

    crw-r----- 1 root sys 307, 1608 date/time /dev/rdsk/ c10t50060E8005271760d5s0 In this example, the major number is 307 and the minor number is 1608. Create an alias device file by executing the following command in each node. An alias device file corresponding to one disk slice should have the same name in all the nodes.

  • Page 95: Memory And Disk Capacity Requirements

    • If the control domain and the I/O domains that manage the HBAs for HDLM-managed storage are not separated: HDLM can be installed in the control domain. ¢ Only domains including I/O domains that manage the HBAs connected to HDLM-managed storage are supported. HDLM cannot be installed in the guest domain.

  • Page 96: Disk Capacity Requirements, The Number Of Paths Supported In Hdlm

    Memory Requirements The host requires 11 MB of memory. Disk Capacity Requirements Table 3-16 Disk capacity requirements for a host on page 3-32 lists the disk capacity requirements for a host. Table 3-16 Disk capacity requirements for a host Directory Disk capacity requirement /etc 2MB + x bytes...

  • Page 97: Flow For Creating An Hdlm Environment

    Table 3-17 Number of LUs and Paths Supported in HDLM Item Supported number Number of LUs 1 to 256 Number of physical paths per 1 to 64 Total number of physical paths 1 to 4096 When Sun Cluster 3.1, Sun Cluster 3.2, or Oracle Solaris Cluster 3.3 is used, configure a server so that the sum of the number of physical paths per an LU for all nodes will be less than or equal to 15 as well as satisfying the conditions shown in...

  • Page 98: Hdlm Installation Types

    Figure 3-1 Flow of HDLM Environment Setup HDLM Installation Types This subsection describes the following types of HDLM installation: new installation, upgrade installation, and re-installation. New installation of HDLM: Installing HDLM in a server on which HDLM has not been installed is called a new installation of HDLM.

  • Page 99: Notes On Creating An Hdlm Environment

    Notes on Creating an HDLM Environment This section gives points to be noted when creating an HDLM environment. For points to be noted during operation of HDLM, see Notes on Using HDLM on page 4-2. Notes on Hardware Settings For details on the settings of the storage system, see storage system manual and maintenance documentation for the storage system.

  • Page 100

    Notes on Installation • You cannot edit the HDLM driver configuration definition file (/kernel/ drv/dlmfdrv.conf). To change an HDLM-managed device to a non- HDLM-managed device, use the HDLM device unconfiguration function of the HDLM device. For details on the HDLM device unconfiguration function, see Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page 3-66...

  • Page 101

    SVS is an HDLM management-target device. ¢ You are upgrading from HDLM 5.6.1 or earlier. ¢ When both these conditions are satisfied, the previous generation of the HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf.bak) and HDLM driver unconfiguration definition file (/ opt/DynamicLinkManager/config/dlmfdrv.unconf.bak) are also updated.

  • Page 102

    Integrated trace files after upgrade installation or re-installation: /var/opt/hitachi/HNTRLib2/spool/hntr2n.log (n indicates a file number) Trace files after upgrade installation or re-installation: /var/opt/DynamicLinkManager/log/hdlmtrn.log (n indicates a file number) • If you interrupted a new installation, check whether the /var/opt/ DynamicLinkManager directory exists. If the /var/opt/ DynamicLinkManager directory exists, execute the following command to delete it: # rm -rf /var/opt/DynamicLinkManager...

  • Page 103

    For details on the commands, refer to the Solaris 10 documentation. Table 3-18 Support for the Solaris zones on page 3-39 shows the HDLM support for the Solaris zones. Table 3-18 Support for the Solaris zones Item Global zone Non-global zone Remarks Installing HDLM Supported...

  • Page 104: Notes On Related Software

    desired kernel mode. If you switch the kernel mode from 32 bits to 64 bits when only the 32-bit version of HDLM is installed, you must reinstall HDLM after the switch is completed. For details on how to switch the kernel mode, see Switch the Kernel Mode on page 3-44 Preparations...

  • Page 105: Notes On Command Execution, Notes On The Disk Label, Installing Hdlm

    You can use zoning of Fibre Channel switches or storage-system's WWN security function to separately use HBA ports. You can also use point-to-point connections to physically separate the HBA ports. Notes on Command Execution When HDLM is used, you cannot specify the -r option in the drvconfig, devlinks, disks, or devfsadm command to reconfigure the HDLM devices on another boot disk.

  • Page 106

    option specified, the logical device files of the sd or ssd devices to be managed by HDLM will be deleted. Back up all pertaining to the logical device files under the /dev/dsk and /dev/rdsk directories to another directory. The following shows an example of executing the commands: # mkdir /var/tmp/bkfile1 # ls -l /dev/dsk >...

  • Page 107: Apply Solaris Patches, Set Up The Hardware

    An example of how to edit the /etc/vfstab file is shown in Figure 3-3 An example of how to edit the /etc/vfstab file on page 3-43. Figure 3-3 An example of how to edit the /etc/vfstab file Comment the shaded portions out by placing a hash mark (#) at the beginning of the line, or delete the line.

  • Page 108: Set Up The /kernel/drv/sd.conf File, Switch The Kernel Mode

    Set Up the /kernel/drv/sd.conf File When using an sd driver, depending on the HBA driver, you need to edit the / kernel/drv/sd.conf file to recognize the disks. For details on how to edit this file, refer to the HBA documentation. When using an ssd driver, you will not need to edit this file, since the connected disks will be automatically recognized.

  • Page 109: Set Up The /etc/system File

    To switch to the 64-bit kernel mode: ¢ If Solaris is running, execute the following command: # reboot -- "boot-device kernel/sparcv9/unix other-boot- options" To boot at the ok prompt, execute the following command: boot boot-device kernel/sparcv9/unix other-boot-options Set Up the /etc/system File HDLM uses a maximum of 8 KB of the stack for the kernel thread.

  • Page 110

    the extended facility. This example shows how to specify the recommended settings in the /etc/syslog.conf file: local0.info /usr/local/audlog Set Up VxVM To use HDLM and VxVM on the same host, set up VxVM as follows: Make sure that the logical device file under the /dev/dsk and /dev/rdsk directories is backed up.

  • Page 111: Set Up Sun Cluster

    Set Up Sun Cluster If you are using Sun Cluster, before installing HDLM, perform the following operations to release the device group and Quorum device settings. • Release the disk device groups that consist of the sd or ssd devices to be managed by HDLM.

  • Page 112

    At another node in the same cluster, switch to become the superuser, and then place the node that was stopped in step 5 into the maintenance mode. The following shows an example of executing the command: host2# /usr/cluster/bin/scconf -c -q node=host1,maintstate Confirm that the cluster node is in the maintenance mode.

  • Page 113: Preparation For Performing An Unattended Installation Of Hdlm

    Preparation for Performing an Unattended Installation of HDLM An unattended installation is an installation in which HDLM is installed based on the responses predefined in the installation-information settings file instead of user responses. The following shows the processing flow of unattended installation: Define information necessary for installing HDLM in the installation- information settings file.

  • Page 114

    Delete this file manually after installation is completed if it is not necessary. Available characters for the absolute path of the license key file are as ¢ follows: Upper case and lower case characters, numbers, and some special characters (/, !, #, +, -, ., =, @, _, ~) Except when performing an unattended installation of HDLM, if a user- ¢...

  • Page 115

    The following shows an example of executing the commands for mounting the DVD-ROM: # mkdir /cdrom/cdrom0 # mount -F hsfs -o ro /vol/dev/dsk/c0t6d0/DVD-ROM-volume-name / cdrom/cdrom0 Perform the installation. If you are not performing an unattended installation of HDLM, execute ¢ either of the following installation commands: When executing the installux.sh script # DVD-ROM-mount-point/installux.sh...

  • Page 116

    saved, is output. Enter 1 if a user-specified license key file has been saved. Enter 2 if a license key is specified directly. Note that if any value other than 1 or 2 is entered three times, the ¢ installation is stopped. If 1 is entered, specify a user-specified license key file with the absolute path.

  • Page 117

    driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf). # /opt/DynamicLinkManager/bin/dlmsetconf Type y in response to the KAPL10242-I message displayed when you execute the dlmsetconf utility. Information pertaining to LUs to which the physical path accesses will be defined in the HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf).

  • Page 118

    boot -r (at the ok prompt) ¢ Execute the dlnkmgr command's view operation to display the status of the programs. The following shows an example of executing the commands: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys HDLM Version : x.x.x-xx Service Pack Version Load Balance : on(extended lio) Support Cluster...

  • Page 119

    # /opt/DynamicLinkManager/bin/dlnkmgr view -path Check the execution results of this command to determine the correspondence between each HDevName (the logical device file name of the HDLM device) and iLU (the internal LU number). Then, compare the correspondence of the logical device file names and internal LUs that took place before HDLM was installed, and then re-register the correct HDLM devices in those programs that will use these devices.

  • Page 120: Performing A New Installation Of Hdlm (when Sun Cluster Is Being Used)

    Item Description Indicates the HDLM 7.2 logical device file name (HDevName). If a NewHDevName hyphen (-) is displayed, the corresponding HDev does not exist after the HDLM 7.2 installation. Indicates the storage system name. DskName Indicates the LU number in the storage system. Based on the results of step 20, compare the logical device file names from before and after the HDLM 7.2 installation, and re-register the correct HDLM devices in programs that use HDLM devices.

  • Page 121

    # format If you want to remove a device from HDLM management, execute the format command to check information for each path of that device. The information you are checking here will be used in step 17. The following shows an example of executing the format command: In this example, the shaded portion shows information for each path of the device you are removing from HDLM management.

  • Page 122

    0x40235c5200000003 In this example, keys are displayed. If keys are displayed, perform the steps beginning with step 7. Even if the Quorum device or disk device group settings are removed, you need to restart the node in non-cluster mode and cancel the key registration because the keys to be used for disk reservation are left registered.

  • Page 123

    If the DVD-ROM cannot be automatically mounted, execute the following command to mount the DVD-ROM: # mount -F hsfs -o ro DVD-ROM-device-name DVD-ROM-mount-point The following shows an example of executing the command for mounting the DVD-ROM: # mkdir /cdrom/cdrom0 # mount -F hsfs -o ro /vol/dev/dsk/c0t6d0/DVD-ROM-volume-name / cdrom/cdrom0 Perform the installation.

  • Page 124

    Note that if any value other than 1 or 2 is entered three times, the ¢ installation is stopped. If 1 is entered, specify a user-specified license key file with the absolute path. Note that if the wrong absolute path is entered three times, the ¢...

  • Page 125

    # /opt/DynamicLinkManager/bin/dlmsetconf Type y in response to the KAPL10242-I message displayed when you execute the dlmsetconf utility. Information associated with LUs that access a physical path is defined in the HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf). Do not edit this file. Specify the devices that are to be excluded from management by HDLM, if there are any.

  • Page 126

    If the nodes were started in cluster mode, execute the following command on any one node: # /usr/cluster/bin/scshutdown -g0 -y With the reconfiguration option specified, start all nodes comprising the cluster in cluster mode, at the same time. Execute boot -r at the ok prompt. Note While a node is starting, the message below may be output to the console or syslog:...

  • Page 127

    By checking the execution results of the view operation, verify the version of the installed HDLM. If HDLM Version is 7.2.0-00, the installed version of HDLM is correct. By checking the execution results of the view operation, verify that the programs are running normally.

  • Page 128

    Item Description Indicates the logical device file name (HDevName) after the HDLM NewHDevName 7.2 upgrade installation. If a hyphen (-) is displayed, the corresponding HDev does not exist after the HDLM 7.2 installation. Indicates the storage system name. DskName Indicates the LU number in the storage system. Based on the results of step 25, compare the logical device file names from before and after the HDLM 7.2 installation, and re-register the correct HDLM devices in programs that use HDLM devices.

  • Page 129

    When the Bourne or Korn shells is used, add the path as shown below: PATH=$PATH:/opt/DynamicLinkManager/bin ; export PATH When the C shell is used, add the path as shown below: set path=( $path /opt/DynamicLinkManager/bin ) Execute the following command to view the HDevName and iLU columns: # /opt/DynamicLinkManager/bin/dlnkmgr view -path Check the execution results of this command to determine the correspondence between each HDevName (logical device file name of the...

  • Page 130

    Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM You can use the HDLM device unconfiguration function to remove Hitachi storage system devices from HDLM management on a per-LU basis. This section explains how to use the HDLM device unconfiguration function. To remove an LU from HDLM management, use a text editor to edit the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/ config/dlmfdrv.unconf).

  • Page 131: Performing An Upgrade Installation Or Re-installation Of Hdlm

    # /opt/DynamicLinkManager/bin/dlmsetconf -u Restart the host with the reconfiguration option specified. Execute one of the following commands: # reboot -- -r ¢ boot -r (at the ok prompt) ¢ Execute the following command to verify that the settings were applied correctly: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv Confirm that the LDEV that was removed from HDLM management is not...

  • Page 132

    immediately after the upgrade installation or re-installation has finished. Do not interrupt the upgrade installation or re-installation. ¢ When you install HDLM while the 32-bit kernel mode is being used, ¢ only the 32-bit version of HDLM will be installed. When you install HDLM while the 64-bit kernel mode is being used, both the 32-bit and 64-bit versions of HDLM will be installed.

  • Page 133

    cdrom/cdrom0 Perform the installation. If you are not performing an unattended installation of HDLM, execute ¢ either of the following installation commands: When executing the installux.sh script # DVD-ROM-mount-point/installux.sh When executing the pkgadd command # pkgadd -a DVD-ROM-mount-point/HDLM_Solaris/DLManager/ install/update -d DVD-ROM-mount-point/HDLM_Solaris DLManager •...

  • Page 134

    If 2 is entered, a message that prompts you to enter a license key is displayed. Enter the license key. When a correct license key is entered, the installation program ¢ displays a message that indicates the license was installed. Take care that if you enter an incorrect license key three times, ¢...

  • Page 135: Installing Hdlm In An Ldoms Environment

    Path Health Checking : on(30) Auto Failback : on(1) Intermittent Error Monitor : off HDLM Manager Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss HDLM Alert Driver Ver WakeupTime ElogMem Size Alive x.x.x-xx yyyy/mm/dd hh:mm:ss 4096 HDLM Driver Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss License Type Expiration Permanent...

  • Page 136

    Figure 3-6 Configuration Example (Before Installation) A virtual disk backend is a location where virtual disk data is stored. A disk, disk slice, file, or volume (such as ZFS, SVM, and VxVM) can be used for the backend. In the control domain, register the install_disk file on a built-in disk as a virtual disk backend.

  • Page 137

    Domain Domain Boot Disk Exporting Virtual Disk Backends Name Type (For a data disk of the guest domain) A guest ldg2 vol1@ldg1-vds0 domain The following figure shows a configuration example for an LDoms environment before installing HDLM. To install HDLM to a host in which LDoms have been set: Log in to the control domain.

  • Page 138

    # ldm stop ldg2 Execute the ldm unbind command to unbind the resource of the guest domain. For example, to unbind a resource of the guest domain ldg2, execute the command as follows: # ldm unbind ldg2 Execute the ldm rm-vdisk command to remove virtual disks from the guest domain.

  • Page 139

    On the control domain, execute the ldm add-vdsdev command to export the virtual disk backends from the service domain on the I/O domain to which HDLM will be installed. For example, to export virtual disk backends vol1 and vol2 from the service domain ldg1, execute the command as follows: # ldm add-vdsdev /dev/dsk/c4t50060E8010027A82d0s2 vol1@ldg1-vds0 # ldm add-vdsdev /dev/dsk/c4t50060E8010027A82d1s2 vol2@ldg1-vds0...

  • Page 140: Configuring A Boot Disk Environment, Overview Of Configuring A Boot Disk Environment

    In this configuration example, /dev/dsk/c4t50060E8010027A82d0s2 and /dev/dsk/c4t50060E8010027A82d1s2 are HDLM devices. Configuring a Boot Disk Environment This section explains how to configure a boot disk environment in which HDLM manages the boot disk. An environment with a boot disk located on the host is called a local boot disk environment.

  • Page 141: Procedure For Configuring A Boot Disk Environment, Migration From An Existing Hdlm Environment

    Perform the procedure described in Migration by Building a New Pre- Migration Environment on page 3-83 Procedure for Configuring a Boot Disk Environment on page 3-77, and then execute the procedure described in Setting Up the Post-Migration Environment on page 3-85 Procedure for Configuring a Boot Disk Environment on page 3-77.

  • Page 142

    Figure 3-8 Flow of setting up the boot disk environment when using an existing HDLM environment as the pre-migration environment Allocate the LUs to be used by the boot disk in the post-migration environment. Get Solaris to recognize the LUs to be used by the post-migration boot disk.

  • Page 143: Migration By Installing Hdlm In The Existing Local Boot Disk Environment

    This step and the next step set the LU to be used for the post-migration boot disk as an HDLM-managed LU. Restart the host with the reconfiguration option specified to make the settings valid and the HDLM device recognizable. Execute one of the following commands: # reboot -- -r ¢...

  • Page 144

    Figure 3-9 Flow of setting up the boot disk environment when installing HDLM in the existing local boot disk environment Allocate the LUs to be used by the boot disk in the post-migration environment. Configure the LU or LUs to be used by the boot disk as follows: To the slices of a single LU, allocate the Solaris file system, swap ¢...

  • Page 145: Migration By Installing Hdlm In The Existing Boot Disk Environment

    Set the LU slices to be used by the post-migration boot disk. Complete the preparations for installing HDLM in the pre-migration environment and then install HDLM, referring to the description in Installing HDLM on page 3-41. Perform the necessary steps. Note Do not exclude the post-migration boot disk from HDLM management when you set the devices to be excluded in step 11 of...

  • Page 146

    Figure 3-10 Flow of setting up the boot disk environment when using an existing boot disk environment without HDLM installed as the pre- migration environment Allocate the LUs to be used by the boot disk in the post-migration environment. Configure the LU or LUs to be used by the boot disk as follows: To the slices of a single LU, allocate the Solaris file system, swap ¢...

  • Page 147: Migration By Building A New Pre-migration Environment

    Get Solaris to recognize the LUs to be used by the post-migration boot disk. Set the LU slices to be used by the post-migration boot disk. Complete the preparations for installing HDLM in the pre-migration environment and then install HDLM, referring to the description in Installing HDLM on page 3-41.

  • Page 148

    Figure 3-11 Flow of setting up the boot disk environment when building a new pre-migration environment Install Solaris in the pre-migration environment. Install Sun Cluster if required. For details, see your Sun Cluster documentation. Install VxVM if required. For details, see your VxVM documentation. Allocate the LUs to be used by the boot disk in the post-migration environment.

  • Page 149: Setting Up The Post-migration Environment

    To the slices of a single LU, allocate the Solaris file system, swap ¢ devices, dump devices, Sun Cluster global device file system, and replica of the SDS or SVM state database. To the slices of multiple LUs, allocate the Solaris file system, swap ¢...

  • Page 150

    # mount /dev/dsk/c8t50060E80004361D1d1s0 /mnt0 # mount /dev/dsk/c8t50060E80004361D1d1s6 /mnt1 # mount /dev/dsk/c8t50060E80004361D1d1s3 /mnt2 Copy the Solaris file system from the pre-migration environment to the post-migration environment. If using Sun Cluster, also copy the Sun Cluster global device file system to the post-migration environment. The following shows an example of executing the command: # ufsdump 0f - / | (cd /mnt0;...

  • Page 151

    /dev/dsk/c8t50060E80004361D1d1s0 /dev/rdsk/ c8t50060E80004361D1d1s0 / ufs 1 no - /dev/dsk/c8t50060E80004361D1d1s6 /dev/rdsk/ c8t50060E80004361D1d1s6 /usr ufs 1 no - /dev/did/dsk/d30s3 /dev/did/rdsk/d30s3 /global/.devices/ node@1 ufs 2 no global Execute the scdidadm -l command to find out the device ID to specify as the mount device of the Sun Cluster global device file system.

  • Page 152

    Configuration Is Changed on page 4-17 Precautions Regarding Changes to the Configuration of an HDLM Operating Environment on page 4-17. To obtain the boot device name specified by OBP, edit the character string of the path determined by the above procedure as follows: - Replace the driver name (ssd or sd) of an ssd or sd driver with disk.

  • Page 153

    Perform steps 13 to 17 only if SDS or SVM is being used and you are ¢ migrating a replica of the SDS or SVM state database to the post- migration environment. In this case, keep the pre-migration environment recognized by Solaris. Skip to step 18 if SDS or SVM is being used but you are not migrating ¢...

  • Page 154: Configuring A Boot Disk Environment For A Zfs File System

    Perform a backup so that you can take action in the following situations: Recover when an error occurs in the post-migration environment. ¢ Boot from the pre-migration environment, and repair the post- migration environment. The post-migration backup can be used to restore the post-migration environment.

  • Page 155: Creating A Zfs Boot Disk Environment

    Figure 3-12 Creating a ZFS boot disk environment Copy the local boot disk environment from the ZFS file system to the LUs (SCSI device) in the storage system. Replace the LUs on the SCSI device in the storage system with those on the HDLM device.

  • Page 156

    • Two LUs have the same capacity as the local boot disk need to be created in the storage system. You also need to configure at least one data disk on the host. • If HDLM has already been installed, remove either of the two LUs created in the storage system from HDLM management.

  • Page 157

    # lustatus Boot Environment Active Active Copy Name Complete Now On Reboot Delete Status --------------------------------------------------------------- s10s_u9wos_14a zfsBE Execute the following commands to allow the system to boot in the ZFS boot disk environment. # luactivate boot-environment-name # installboot -F zfs /usr/platform/`uname -i`/lib/fs/zfs/ bootblk LU-SCSI-device-(raw-device) Note that the SCSI device specified for the installboot command must be a raw device.

  • Page 158

    Replacing the ZFS boot disk environment on the SCSI device with the ZFS boot disk environment on the HDLM device Set up a device for the ZFS boot disk environment created in Boot Disk Environment that uses a ZFS File System on page 3-90 on the HDLM- managed device.

  • Page 159: Migrating From A Zfs Boot Disk Environment To The Local Boot Disk Environment

    Execute the installboot command to set the LU#1 to be used for a ZFS boot disk environment. Execute the following command to allow the system to boot in a new boot disk environment. # installboot -F zfs /usr/platform/`uname -i`/lib/fs/zfs/ bootblk HDLM-device-(raw-device)-after-replacement Note that the SCSI device specified for the installboot command must be a raw device.

  • Page 160: Migrating From A Boot Disk Environment To The Local Boot Disk Environment

    # luactivate boot-environment-name # installboot -F zfs /usr/platform/`uname -i`/lib/fs/zfs/ bootblk internal-disk-device-(raw-device) Note that the device on the built-in disk specified for the installboot command must be a raw device. In the following example, the boot disk environment name is zfsBE2 and the name of the device on the built-in disk is c1t0d0s0: # luactivate zfsBE2 # installboot -F zfs /usr/platform/`uname -i`/lib/fs/zfs/...

  • Page 161

    This method migrates to the local boot disk environment with the current operating status intact. The procedure is outlined and described below. An overview of the procedure is shown in Figure 3-13 Migration flow when coping the file system from a boot disk environment to the local boot disk environment on page 3-97.

  • Page 162

    Get Solaris to recognize the LUs to be used by the post-migration boot disk. Set the LU slices to be used by the post-migration boot disk. If using Sun Cluster, go to the next step. ¢ If not using Sun Cluster, go to step 8. ¢...

  • Page 163

    Copy the Solaris file system from the pre-migration environment to the post-migration environment. If using Sun Cluster, also copy the Sun Cluster global device file system to the post-migration environment. The following shows an example of executing the command: # ufsdump 0f - / | (cd /mnt0; ufsrestore rf -) # ufsdump 0f - /usr | (cd /mnt1;...

  • Page 164

    After editing /dev/dsk/c0t0d0s1 - - swap - no - Edit the post-migration boot disk's /etc/dumpadm.conf file, changing the dump device definition in the DUMPADM_DEVICE parameter from the pre- migration boot disk to the post-migration boot disk. An example is shown below. Before editing DUMPADM_DEVICE=/dev/dsk/c8t50060E80004361D1d1s1 After editing...

  • Page 165

    Perform the required settings if you want to mirror or multi-path the post- migration boot disk in the following applications. For details on the required settings, see the documentation for the relevant application. ¢ ¢ VxVM ¢ Sun StorageTek Traffic Manager software (or, in Solaris 10, the ¢...

  • Page 166: Mirroring The Environment By Using Svm, From The Local Boot Disk Environment

    • For details about the action to take if there is a path error in the boot disk environment, see Actions to Take for a Path Error in a Boot Disk Environment on page 5-6. • In the boot disk environment, Solaris cannot be upgraded. If Solaris needs to be upgraded, perform the following: Uninstall HDLM after migration to a local boot disk environment by performing the procedures in...

  • Page 167

    Figure 3-14 Configuring a boot disk environment in which HDLM manages the boot disk, from the local boot disk environment Mirroring a Boot Disk Environment in Which HDLM Manages the Boot Disk by Using SVM Mirror a configured boot disk by using SVM. Figure 3-15 Mirroring a boot disk environment in which HDLM manages the boot disk on page 3-103 shows...

  • Page 168

    • Slice 0: / (root) file system • Slice 1: Swap device and dump device • Slice 6: /usr file system • Slice 7: SVM state database Prepare another LU with the same capacity as that of the created boot disk, and make the configuration and capacity of slices the same.

  • Page 169

    Execute the metaroot command. # metaroot d10 Reboot the host. When the host starts, check the swap device and dump device. Then execute the mount command to confirm that / (root) and /usr have been correctly configured. # swap –l swapfile swaplo blocks free...

  • Page 170: Boot Disk Environment Incorporating Svm

    Stripe 0: Device Start Block Dbase State Reloc Hot Spare /dev/dsk/ c8t50060E08000436D1d4s0 Okay d12: Submirror of d10 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Start Block Dbase State Reloc Hot Spare /dev/dsk/ c8t50060E08000436D1d5s0 Okay Execute the dlnkmgr command's view -path operation to confirm that an I/O is issued to the LU set as the boot disk.

  • Page 171: Installing Hdlm And Then Configuring The Environment, Placing The Boot Disks Under Hdlm Management

    Installing HDLM and then Configuring the Environment Install HDLM by referring to Installing HDLM on page 3-41. Execute the dlmsetconf utility for creating the HDLM driver configuration definition file to configure the HDLM device. Remove the LUs of the mirrored boot disks from HDLM management. For details, see Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page...

  • Page 172

    • Slice 7: SVM state database Decide the metadevice names. The metadevice names of the current boot disks and the boot disks after the migration are specified as shown in Table 3-22 Metadevice names on page 3-108. Table 3-22 Metadevice names Boot disk File system Mirror...

  • Page 173

    # newfs /dev/md/rdsk/d300 Mount the Solaris basic file system of the post-migration environment. For example: # mount /dev/md/dsk/d100 /mnt0 # mount /dev/md/dsk/d300 /mnt1 Copy the existing Solaris basic file system to the post-migration environment. For example: # ufsdump 0f - / | (cd /mnt0; ufsrestore rf -) # ufsdump 0f - /usr | (cd /mnt1;...

  • Page 174

    To determine this path, perform the procedure in When the Path Configuration Is Changed on page 4-17 Precautions Regarding Changes to the Configuration of an HDLM Operating Environment on page 4-17. To obtain the boot device name specified by OBP, edit the character string of the determined path as follows: Replace the driver name (ssd) of an ssd driver with disk.

  • Page 175

    # swap -l swapfile swaplo blocks free /dev/md/dsk/d200 85,200 16 8409584 8409584 # dumpadm Dump content: kernel pages Dump device: /dev/md/dsk/d200 (swap) Savecore directory: /var/crash/host Savecore enabled: yes # mount / on /dev/md/dsk/d100 read/write/setuid/devices/intr/largefiles/ logging/xattr/onerror=panic/dev=1540064 on Mon Mar 9 11:00:25 2009 /usr on /dev/md/dsk/d300 read/write/setuid/devices/intr/ largefiles/logging/xattr/onerror=panic/dev=154012c on Mon Mar...

  • Page 176

    Size: 25190400 blocks (12 GB) d101: Submirror of d100 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Start Block Dbase State Reloc Hot Spare /dev/dsk/c8t50060E08000436D1d6s0 Okay d102: Submirror of d100 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Start Block Dbase...

  • Page 177: Uninstalling Hdlm, Excluding The Prepared Lus From Hdlm Management

    Backups of the environments are required for the following situations: If an error occurs in the post-migration environment: ¢ Reboot from the pre-migration environment, and then use it to restore the post-migration environment. Alternatively, the backup of the post-migration environment can also be restored. If you want to migrate from a boot disk environment to a local boot ¢...

  • Page 178

    Figure 3-18 Current boot disks and boot disks after the migration The LU device names of the current boot disks are c8t50060E08000436D1d4 and c8t50060E08000436D1d5, and the LU device names of the boot disk after the migration are c6t50060E08000436D1d6 and c6t50060E08000436D1d7, as shown in the above figure.

  • Page 179

    Boot disk File system Mirror Sub mirror /usr d300 d301 d302 After migration /(root) swap /usr Create metadevices on the sd/ssd device, and then mirror them. (Single mirror) # metainit -f d11 1 1 c6t50060E08000436D1d6s0 # metainit d10 -m d11 # metainit -f d21 1 1 c6t50060E08000436D1d6s1 # metainit d20 -m d21 # metainit -f d31 1 1 c6t50060E08000436D1d6s6...

  • Page 180

    # ufsdump 0f - /usr | (cd /mnt1; ufsrestore rf -) Edit the file /etc/vfstab (/mnt0/etc/vfstab) of the boot disk after the migration to change the definition of the swap after the migration to a metadevice (/dev/md/dsk/d20) and change the definition of /usr after the migration to a metadevice (/dev/md/ds/d30).

  • Page 181

    Figure 3-19 Booting from the post-migration environment When the host starts, confirm that the swap device and dump device have been correctly configured. Then, use the mount command to confirm that / (root) and /usr have been correctly configured. # swap -l swapfile swaplo blocks free...

  • Page 182

    # metainit -f d22 1 1 c6t50060E08000436D1d7s1 # metattach d20 d22 # metainit -f d32 1 1 c6t50060E08000436D1d7s6 # metattach d30 d32 Copy the SVM state database to the boot disk after migration.. # metadb -a c6t50060E08000436D1d6s7 # metadb -a c6t50060E08000436D1d7s7 Delete the SVM in the pre-migration environment.

  • Page 183: Checking The Path Configuration

    Uninstall HDLM by referring to Uninstalling HDLM on page 3-164. Checking the Path Configuration HDLM functions, such as load balancing and failover, are only available for HDLM management-target devices that have more than one active path. After you install HDLM or change the hardware configuration, check the structure and statuses of the paths.

  • Page 184: Setting Up Hdlm Functions, Checking The Current Settings, Setting Up The Hdlm Functions

    Setting Up HDLM Functions HDLM includes functions like the load balancing function, the automatic failback function, the error logging function, the audit logging function, etc. You can set up these functions by using the dlnkmgr command's set operation. The following subsections describe these setup methods. Checking the Current Settings Before setting the HDLM functions by using the dlnkmgr command's set operation, check the current settings by executing the following command:...

  • Page 185: Setting Up Load Balancing

    Function Default value Recommended value Path health checking 30-minute check interval The recommended checking interval depends on the operating environment. Automatic failback The recommended checking interval depends on the 1-minute check interval operating environment. Intermittent Error Monitor Logging level 3: Collects all the error 3: Collects all the error information for the information for the...

  • Page 186: Setting Up Path Health Checking, Setting Up The Automatic Failback Function

    When you set this value to on, specify rr for round robin or exrr for extended round robin after the -lbtype option. The type of algorithm specified by the -lbtype parameter remains stored in the system, even when you disable the load balancing function by specifying -lb off. Therefore, when you re-enable the load balancing function without specifying an algorithm, load balancing will be executed according to the algorithm stored in the system.

  • Page 187: Setting Up Intermittent Error Monitoring, Setting The Error Log Collection Level

    be applied. For example, specify the auto failback as off after specifying the checking interval to five minutes and executing. Then, when executing after specifying the auto failback as on without specifying the checking interval, the five minutes that were specified previously will be applied again. Setting Up Intermittent Error Monitoring You can set the intermittent error monitoring only when automatic failback is enabled.

  • Page 188: Setting The Trace Level

    Table 3-25 Values for the error log collection level setting Value Description No error logs are collected. All information for errors of the "Error" level or higher is collected. All information for errors of the "Warning" level or higher is collected. All information for errors of the "Information"...

  • Page 189: Setting The Error Log File Size, Setting The Number Of Error Log Files

    For normal operation, we recommend that you set the trace level to 0. If you set the trace level to a value higher than necessary, HDLM performance might decrease, or trace information required to analyze the cause of an error might be overwritten. The following is an example of setting up the trace level by using a command: # /opt/DynamicLinkManager/bin/dlnkmgr set -systflv 1...

  • Page 190: Setting The Trace File Size, Setting The Number Of Trace Files

    Setting the Trace File Size Trace files for which you can set a trace file size are hdlmtrn.log (n indicates a file number from 1 to 64). The length of a trace file is fixed, regardless of how much trace information is actually in the file. For the trace file size, you can specify a value (in kilobytes) from 100 to 16000.

  • Page 191: Setting The Audit Log Facility

    Table 3-27 Values Indicating Audit Log Data Collection Levels Value (severity) Explanation No audit log data is collected. Critical-level audit log data is collected. Critical-level and Error-level audit log data is collected. Critical-level, Error-level, and Warning-level audit log data is collected.

  • Page 192: Checking The Updated Settings

    define the output destination of the facility in the /etc/syslog.conf file. Table 3-29 Values Indicating Audit Log Facility on page 3-128 describes the values that indicate audit log facility. The default is user. Table 3-29 Values Indicating Audit Log Facility Value Corresponding facility value in the /etc/syslog.conf file user or 1...

  • Page 193: Setting Up Integrated Traces

    After you have set up the collection of audit log data, use the following command to make sure that the setting has been specified correctly: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -audlog Audit Log : on(6) Audit Log Category : all Audit Log Facility : local0 KAPL01001-I The HDLM command completed normally.

  • Page 194: Notes On Using The Hitachi Network Objectplaza Trace Library

    mind that programs other than HDLM also use them. If an earlier Hitachi Network Objectplaza Trace Library (HNTRLib) is already installed, the settings will not be inherited. Notes on Using the Hitachi Network Objectplaza Trace Library Note the following when using Hitachi Network Objectplaza trace library: •...

  • Page 195: Changing The Size Of Integrated Trace Files, Changing The Number Of Integrated Trace Files

    Changing the Size of Integrated Trace Files To change the size of integrated trace files: In the Hitachi Network Objectplaza Trace Library setup menu, type 1 and then press the Enter key. A screen to set the size of the integrated trace file appears. The current value is displayed in Current Size(KB).

  • Page 196: Changing The Buffer Size Per Monitoring Interval Duration

    Press the Enter key. The new setting is applied and the Hitachi Network Objectplaza Trace Library setup menu appears again. Changing the Buffer Size Per Monitoring Interval Duration To change the buffer size per monitoring interval: In the Hitachi Network Objectplaza Trace Library setup menu, type 4, and then press the Enter key.

  • Page 197: Adjusting The Number Of Messages To Be Output Per Monitoring Interval

    Press the Enter key. The new setting is applied and the Hitachi Network Objectplaza Trace Library setup menu appears again. Adjusting the Number of Messages to Be Output Per Monitoring Interval To adjust the number of messages to be output to fit a particular monitoring interval: In the Hitachi Network Objectplaza Trace Library setup menu, type 6, and then press the Enter key.

  • Page 198: Finishing The Hitachi Network Objectplaza Trace Library Settings

    Adjust the maximum number of messages output to the integrated trace files in New Max(sec). The default is 0. You can specify a value from 0 to 500 (messages). If you want to increase the number of messages that are output to the integrated trace file as much as possible, we recommend that you set this value to 0.

  • Page 199

    If you do not know how to stop and start programs other than HDLM, do not perform the following steps, and restart the host. Execute the following command to stop the integrated trace collection process: # /opt/hitachi/HNTRLib2/bin/hntr2kill Execute the following command to delete the memory mapped file: # rm /opt/hitachi/HNTRLib2/mmap/hntr2mmap.mm Execute the following command to start the integrated trace collection process:...

  • Page 200: Setting Up Vxvm

    Execute a command such as the following: # mount -F ufs /dev/dsk/c2t7d16s0 mount-point Enable automatic mounting. To automatically mount the HDLM-managed device when the host starts, edit the /etc/vfstab file. Execute the following command to display the current setting: # mount -p Figure 3-20 An output example for the mount -p command on page 3-136 shows an output example of the command.

  • Page 201: Creating A Disk Group

    Creating a Disk Group This subsection mainly describes the preparations for registering the HDLM device in the VxVM disk group. To register the HDLM device in the VxVM disk group, you must remove non- HDLM devices from VxVM. If you have not yet removed non-HDLM devices, first remove them as described in Removing Devices from VxVM on a Controller Basis on page 3-142...

  • Page 202

    If a device name of an enclosure (HDS9980V0_1) is displayed in the DEVICE column for the device, execute the following command to check the device name, in the format cUtXdYsZ, corresponding to the device name of the enclosure. Note The device name of the enclosure corresponding to the LU might vary before and after installing HDLM.

  • Page 203: Creating Vxvm Volumes

    displayed for HDevName of the dlnkmgr command's view -path operation), and if STATE is ENABLED. Therefore, non-HDLM devices must be removed from VxVM management. To remove non- HDLM devices from VxVM management, see Removing Devices from VxVM on a Controller Basis on page 3-142 Removing Devices From VxVM on a Path Basis on page 3-146 Removing Devices from VxVM on...

  • Page 204: Removing Devices From Vxvm, Devices To Be Removed From Vxvm

    Removing Devices from VxVM VxVM uses HDLM devices to access LUs managed by HDLM. Therefore, you must remove non-HDLM devices (such as sd or ssd devices), which correspond to the LU, from VxVM management. In this section, excluding devices from VxVM management is called removing devices.

  • Page 205

    • sd or ssd devices (c3t0d0, c4t1d0) When an HDLM device is created, the logical device file of the corresponding sd or ssd device will be deleted. However, VxVM recognizes the sd or ssd device as still existing even if the logical device file was deleted.

  • Page 206: Removing Devices From Vxvm On A Controller Basis

    The settings files necessary for device suppression are created by selecting menu task 17 (Prevent multipathing/Suppress devices from VxVM's view) or 18 (Allow multipathing/Unsuppress devices from VxVM's view) from the vxdiskadm main menu, and then by selecting menu task 8 (List currently suppressed/non-multipathed devices) from the VolumeManager/Disk/ExcludeDevices or VolumeManager/Disk/IncludeDevices menu.

  • Page 207

    Figure 3-24 When HBA ports on the alternate path differ from those on the primary path connected to the Hitachi storage system To remove devices from VxVM, in the above configuration: Use the dlmvxexclude utility to acquire the mapping between controller numbers and devices.

  • Page 208

    The shaded portion, in the next step, is contained in the /etc/vx/ vxvm.exclude file. Make an entry for the controller of the device you wish to remove from VxVM in the controllers section of the /etc/vx/vxvm.exclude file. In the /etc/vx/vxvm.exclude file, write the shaded portion of the execution result of the dlmvxexclude utility shown in step 1.

  • Page 209

    The following shows an example of executing the command: The devices displayed for HDevName (the shaded portion) are the HDLM devices that must be recognized by VxVM. In the above example, two devices, c5t0d0 and c5t0d1, must be recognized. Execute the vxdisk command to list devices that are recognized by VxVM.

  • Page 210: Removing Devices From Vxvm On A Path Basis

    If any non-HDLM devices are ENABLED, perform the procedure described Actions To Be Taken if an sd or ssd Device Has Not Been Suppressed from VxVM on page 3-150 Removing Devices from VxVM on page 3-140 to set up vxvm.exclude again. When the execution result of the vxdmpadm command is the status shown Table 3-31 When setup of vxvm.exclude is unnecessary on page 3-146,...

  • Page 211

    Figure 3-26 When the HBA Port of the Physical Path Connected to a Hitachi Storage System Is the Same As the One Connected to a Third-party Storage System To remove devices from VxVM, in the above configuration: Use the dlmvxexclude utility to acquire the mapping between controller numbers and devices.

  • Page 212

    The shaded portion indicates the devices that need to be removed from VxVM. Make an entry for the device you wish to removed from VxVM in the paths section of the /etc/vx/vxvm.exclude file. In the /etc/vx/vxvm.exclude file, write the shaded portion of the execution result of the dlmvxexclude utility shown in step 1.

  • Page 213

    The devices displayed for HDevName (the shaded portion) are the HDLM devices that must be recognized by VxVM. In the above example, it is c4t0d0. Execute the vxdisk command to list devices that are recognized by VxVM. The following shows an example of executing the command: If the HDLM devices output by the dlnkmgr command's view -path operation (the shaded portion) are displayed, the HDLM devices have been recognized by VxVM.

  • Page 214

    from VxVM on page 3-150 Removing Devices from VxVM on page 3-140 to set up vxvm.exclude again. When the execution result of the vxdmpadm command is the status shown Table 3-32 When setup of vxvm.exclude is unnecessary on page 3-150, you do not need to set up vxvm.exclude.

  • Page 215

    Example 1: The vxdisk command does not display HDLM devices but displays the sd or ssd devices corresponding to those HDLM devices. The vxdisk command displays the SCSI device c3t0d0, which corresponds to an HDLM device, instead of the HDLM device c5t0d0. Example 2: The vxdmpadm command outputs a non-HDLM device, and the STATE column shows ENABLED.

  • Page 216

    The vxdmpadm command displays c3t0d0 and c4t1d0 in addition to the HDLM device c5t0d0, and the STATE column shows ENABLED. If sd or ssd devices have not been removed normally from VxVM, as shown in Examples 1 and 2, take the following action: To suppress sd or ssd devices: Execute the vxdmpadm command.

  • Page 217

    Figure 3-28 The /etc/vx/vxvm.exclude file (before editing) Figure 3-29 The /etc/vx/vxvm.exclude file (after editing) Among the non-HDLM devices in the ENABLED state, there might be two or more devices whose tXdY portions in the logical device file name (cUtXdYs2 format) are the same. In this case, change the currently assigned controller number in /etc/vx/vxvm.exclude.

  • Page 218: Linking Vxvm And Sun Cluster

    If the STATE column shows DISABLED, the device has been removed from VxVM. The following shows an example of executing the command: If there is more than one candidate for the controller number to be assigned in step 4, you might not be able to remove devices from VxVM in one operation.

  • Page 219

    Verify that the VxVM disk group has been imported into only one of the nodes that constitute a cluster. If the VxVM disk group has been deported from all nodes, import the VxVM disk group to be registered in Sun Cluster into only one of the nodes.

  • Page 220

    KAPL05023-E Could not process the IOCTL(00004D03). (0000000D:00000000) If the VxVM volume has not been created, create a VxVM volume in the disk group from the master node (the node that currently acts as the master for the disk device group). The following shows an example of executing the command: # vxassist -g dlmdg make vol-dg 10g To create a volume or otherwise modify the VxVM disk group after it has...

  • Page 221: Registering Hdlm Devices, To Use A Local Metadevice

    • Even if you use a local diskset or shared diskset, do not specify the -r option when changing the path configuration and executing the dlmsetconf utility for creating the HDLM driver configuration definition file. When you specify the -r option, the minor number of the HDLM device used on the SDS will be changed, and you will no longer be able to use the metadevice.

  • Page 222: To Use A Shared Diskset

    host1# metadb -a -f c1t1d0s1 c1t1d0s3 c1t1d0s7 Create a local metadevice. Use the HDLM device (slice) to create a local metadevice (logical disk). The types of metadevices include coupled, stripe, mirrored, RAID5, and transmeta devices. The following shows an example of executing the command to create local metadevice (d0) of type stripe (1) comprised of two HDLM devices (slices).

  • Page 223

    host1# metaset -s diskset1 -a -h host1 host2 Register a disk into the shared diskset. The following shows examples of executing the commands: When not using Sun Cluster: host1# metaset -s diskset1 -a c5t1d2 c5t1d3 When using Sun Cluster: host1# metaset -s diskset1 -a /dev/did/dsk/d10 /dev/did/dsk/ Create a shared metadevice in the shared diskset.

  • Page 224

    Setting Up SVM If SVM is installed on an HDLM host, SVM uses the logical device file name of the HDLM device to access the HDLM-managed device. Therefore, after installing HDLM, you need to register the logical device file name of the HDLM device in SVM.

  • Page 225: To Use A Local Volume, Registering Hdlm Devices, To Use A Shared Diskset

    Registering HDLM Devices If SVM is used, perform the following settings after installing HDLM, and register the SVM volume. To Use a Local Volume To use a local volume: Create a replica of the state database. Create three replicas of the local state database on the disk (slice). Create these replicas on a disk that is not managed by HDLM, not shared, and has not been connected to a Fibre Channel.

  • Page 226

    Create these replicas on all hosts that use the shared diskset. Make sure that all the following conditions are satisfied for the creation-target disk: The disk is not managed by HDLM. ¢ The disk is not shared. ¢ The disk is not connected via Fibre Channel. ¢...

  • Page 227

    host1# mkdir /diskset1 host1# mount /dev/md/diskset1/dsk/d1 /diskset1 When using Sun Cluster: host1# mkdir -p /global/diskset1/d1 host2# mkdir -p /global/diskset1/d1 host1# mount -g /dev/md/diskset1/dsk/d1 /global/diskset1/d1 Setting Up VCS When VCS is installed on an HDLM host, you can register HDLM devices, or VxVM disk groups containing HDLM devices, as resource groups.

  • Page 228: Overview Of Hdlm Uninstallation

    b: The number of paths for each HDLM device c: The I/O timeout value for the sd or ssd driver d: The total number of all HDLM devices If the number of paths differs depending on an HDLM device, use the maximum number for the calculation.

  • Page 229: Preparations For Hdlm Uninstallation, Performing Operations On Hdlm-managed Devices

    Preparations for HDLM Uninstallation Performing Operations on HDLM-Managed Devices To perform operations on HDLM-managed disks: If necessary, back up the contents of the HDLM-managed devices to tape or some other storage medium. For the HDLM devices, collect all relevant information on the correspondence between the logical device file names and the internal LUNs of the LUs in the storage system.

  • Page 230: Remove Sun Cluster Settings

    Use the following command to unmount all mount points that are using HDLM: # umount mount-point If HDLM devices are being used in Solaris 10 ZFS, execute the following command to export all ZFS storage pools that are using HDLM: # zpool export pool-name Edit the /etc/vfstab file.

  • Page 231

    To unregister a VxVM disk device group, enter 3 (Unregister a VxVM device group). As the utility instructs, enter the name of the VxVM disk device group you want to unregister. If the HDLM device has been configured as a Quorum device, perform the following operations to remove the Quorum device settings.

  • Page 232

    Confirm that the cluster node enters the maintenance mode. Make sure that the node that is in the maintenance mode has the following status: The node is in the Offline status. ¢ Both the current Quorum votes present and Quorum votes ¢...

  • Page 233: Remove Vcs Settings, Remove Vxvm Settings, Remove Sds Settings

    Remove VCS Settings If you are using VCS, before uninstalling HDLM, unregister the HDLM devices and VxVM disk groups containing HDLM devices, from the resource groups. For details on how to remove the VCS settings, refer to the VCS documentation. Remove VxVM Settings If you are using both HDLM and VxVM on the same host, before you uninstall HDLM, perform the following to delete the HDLM devices from VxVM.

  • Page 234: Remove Svm Settings

    Note Deleting a metadevice deletes data in the metadevice. Make sure that you first back up the data. When using a local metadevice Execute the following command to unmount the file system The following shows an example of executing the command: host1# umount /mnt/filesystem1 Delete the local metadevice that is using the HDLM device.

  • Page 235: Uninstalling Hdlm From The Local Boot Disk Environment, Uninstalling Hdlm

    The following shows an example of executing the command: host1# umount /mnt/filesystem1 Delete the SVM local volume that is using the HDLM device. The following shows an example of executing the command to delete SVM local volume d0. host1# metaclear d0 When using a shared diskset Delete the shared volume.

  • Page 236

    To uninstall HDLM: Log in to Solaris as a user with root permissions. Execute the following command to uninstall HDLM: # pkgrm DLManager Type y in response to the following message: The following package is currently installed: DLManager Dynamic Link Manager (sparc) xx.x.x.xxxx Do you want to remove this package? [y,n,?,q] Type y in response to the following message:...

  • Page 237: Uninstalling Hdlm From The Boot Disk Environment, Uninstalling Hdlm From An Ldoms Environment

    This message is output because HDLM is uninstalled and the HDLM devices are deleted. If you do not want this message to be output next time the node starts, update the device ID information by referring to steps 1 to 4 in Sun Cluster Settings on page 3-178 Settings Needed After HDLM Uninstallation on page 3-178.

  • Page 238

    In this configuration example, /dev/dsk/c4t50060E8010027A82d0s2 and / dev/dsk/c4t50060E8010027A82d1s2 are HDLM devices. An LDoms domain setting example is shown in Table 3-34 LDoms Setting Example on page 3-174. Table 3-34 LDoms Setting Example Domain Domain Boot Disk Exporting Virtual Disk Backends Name Type A control...

  • Page 239

    To uninstall HDLM from a host in which LDoms have been set: Log in to the control domain. Execute the ldm stop command to stop the guest domain. For example, to stop a guest domain ldg2, execute the command as follows: # ldm stop ldg2 Execute the ldm unbind command to unbind the resource of the guest...

  • Page 240

    For example, to remove virtual disks vdisk1 and vdisk2 from the guest domain ldg2, execute the command as follows: # ldm rm-vdisk vdisk1 ldg2 # ldm rm-vdisk vdisk2 ldg2 Execute the ldm rm-vdsdev command to stop exporting virtual disk backends from the service domain on the I/O domain to which HDLM will be installed.

  • Page 241

    On the control domain, execute the ldm add-vdisk command to assign the virtual disk backends to the guest domain. For example, to assign the virtual disk backends vol1 and vol2 exported from the service domain ldg1 to the guest domain ldg2, execute the command as follows: # ldm add-vdisk vdisk1 vol1@ldg1-vds0 ldg2 # ldm add-vdisk vdisk2 vol2@ldg1-vds0 ldg2...

  • Page 242: Sun Cluster Settings

    Settings Needed After HDLM Uninstallation VxVM Settings When using VxVM, execute the following command to confirm that the DEVICE column of the LUs previously managed by HDLM is displayed in the cWtXdYsZ format (for example, c6t0d10s2): # vxdisk list If the value is displayed in the cWtXdYsZ format, perform the following procedures: Delete the /etc/vx/disk.info file.

  • Page 243

    Remove the device ID information of devices that do not exist. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -C Assign device IDs. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -r Notify the server of the new device ID information. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -ui Update the Sun Cluster global device namespace.

  • Page 244: File System Settings, Application Program Settings, Uninstalling Hitachi Network Objectplaza Trace Library (hntrlib2)

    File System Settings For the file systems of the LUs that were managed by HDLM, specify settings that allow the Solaris sd or ssd driver to be mounted or specify settings that allow volume manager's volumes to be mounted after setup is performed again.

  • Page 245: Uninstalling Hitachi Network Objectplaza Trace Library (hntrlib)

    If you are unable to complete uninstallation even though no programs other than HDLM are using HNTRLib2, contact your HDLM vendor or maintenance company if there is a maintenance contract for HDLM. Note If the log output directory set in HNTRLib2 was not the default directory, the log files will not be deleted during uninstallation.

  • Page 246

    Even if you attempt to uninstall HNTRLib2 when HNTRLib is installed, ¢ HNTRLib will not be uninstalled. If no other programs are using HNTRLib, delete it manually. 3-182 Creating an HDLM Environment Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 247: Hdlm Operation

    HDLM Operation This chapter describes operating procedures for HDLM, including how to operate HDLM and the HDLM manager, and how to change the configuration of the operating environment. □ Notes on Using HDLM □ HDLM Operations Using Commands □ Starting and Stopping the HDLM Manager □...

  • Page 248: When A Path Error Is Detected, Iostat Command

    Notes on Using HDLM This section explains the notes on using HDLM and using the environment in which HDLM is installed. Make sure that you read this section. Displaying Path Information • The path IDs displayed by using the dlnkmgr command's view operation differ depending on the order in which paths were detected when the host started.

  • Page 249: Starting The Host, System, And Hdlm In A Solaris 10 Environment

    The shaded portion indicates mapping from combinations of sd or ssd device types and instance numbers to HDLM device logical device file names. Storage System • Start up the storage system before you start up the hosts, so that Solaris can detect the storage system.

  • Page 250: Upgrading Solaris, Operation In Single-user Mode

    If the host is started in single-user mode, the displayed path status indicates an error even when the actual path status is normal. Therefore, before you use HDLM devices in single-user mode, execute the procedure described in Initializing HDLM When the Host Is Started in Single-User Mode on page 4-4Operation in Single-User Mode on page 4-4.

  • Page 251: Tasks That Can Be Performed In Single-user Mode

    wait at least 10 seconds and then re-execute the command. If still unsuccessful, take the action indicated in the displayed message. Execute the following command to set the HDLM logical device file name in the HDLM filter driver: # /opt/DynamicLinkManager/bin/dlmsetdevname Execute the following command to start the HDLM manager: # /etc/init.d/DLMManager start Tasks that Can Be Performed in Single-User Mode...

  • Page 252: Maintenance Tasks On Devices Connected By Paths In The Boot Disk Environment

    Maintenance Tasks on Devices Connected by Paths in the Boot Disk Environment Note the following points when performing maintenance on devices connected to the boot disk in a boot disk environment running Solaris: • Ensure that at least one path connected to the boot disk is online when you perform a maintenance task.

  • Page 253: Changing The Status Of Paths, Changing The Status Of Paths To Online

    For details on the displayed items and their descriptions, see view (Displays Information) on page 6-30. Changing the Status of Paths This section explains how to change path statuses. Changing the Status of Paths to Online To change the status of paths to online: Check the current status of the paths.

  • Page 254: Changing The Status Of Paths To Offline(c)

    KAPL01061-I 3 path(s) were successfully placed Online; 0 path(s) were not. Operation name = online Check to see if the statuses of all the applicable paths have changed. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -path Changing the Status of Paths to Offline(C) To change the status of paths to Offline(C): Check the current status of the paths.

  • Page 255: Viewing Lu Information

    # /opt/DynamicLinkManager/bin/dlnkmgr view -path Viewing LU Information This section explains how to display LU information by using the dlnkmgr command. To display LU information, execute the dlnkmgr command's view operation with the -lu parameter specified. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -lu Product...

  • Page 256: Initializing Statistical Information For Paths

    The following shows an example of executing the view operation: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv PathID HDevName Device LDEV 000000 c6t5d0 sd74 9970/9980.15001.0670 000001 c6t5d1 sd75 9970/9980.15001.0671 000002 c6t5d2 sd76 9970/9980.15001.0672 000003 c6t5d3 sd77 9970/9980.15001.0673 000004 c6t5d4 sd78 9970/9980.15001.0674 000005 c6t5d5 sd79 9970/9980.15001.0675 000006 c6t5d0...

  • Page 257: Viewing And Setting Up The Operating Environment, Viewing The Operating Environment

    # /opt/DynamicLinkManager/bin/dlnkmgr view -path Viewing and Setting Up the Operating Environment This section explains how to display and set up the HDLM operating environment. Viewing the Operating Environment To display the operating environment, execute the dlnkmgr command's view operation with the -sys and -sfunc parameters specified. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -sfunc HDLM Version...

  • Page 258: Viewing License Information

    • Path health checking • Automatic failback • Intermittent error monitoring • Error log collection level • Trace level • Error log file size • The number of error log files • Trace file size • The number of trace files •...

  • Page 259: Updating The License, Viewing Hdlm Version Information

    Updating the License This section explains how to update the license. To update the license, execute the dlnkmgr command's set operation with the -lic parameter specified. When the confirmation message is displayed, enter y to execute, or n to cancel the command. If the license key file does not exist, a message asking you to enter the license key appears, so enter the license key.

  • Page 260: Viewing Hdlm Component Information

    Alive x.x.x-xx yyyy/mm/dd hh:mm:ss HDLM Alert Driver Ver WakeupTime ElogMem Size Alive x.x.x-xx yyyy/mm/dd hh:mm:ss 4096 HDLM Driver Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss License Type Expiration Permanent KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss The value displayed in HDLM version indicates the HDLM version.

  • Page 261: Starting And Stopping The Hdlm Manager, Starting The Hdlm Manager, Stopping The Hdlm Manager

    # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -pdrv Starting and Stopping the HDLM Manager If an error occurs in the system, such as in an HDLM program, you may need to manually stop or start the HDLM manager to recover from the error. Starting the HDLM Manager To start the HDLM manager, log in to Solaris as a user with root permissions and then execute the following command.

  • Page 262: Hdlm Resident Processes, Changing The Configuration Of The Hdlm Operating Environment

    view, completion time = yyyy/mm/dd hh:mm:ss When the HDLM Manager column shows Dead, the HDLM manager is inactive. HDLM Resident Processes Table 4-1 HDLM resident processes on page 4-16 lists and describes the resident processes in HDLM. To monitor these processes, use the names below.

  • Page 263: Environment, Changing The Configuration Of A System That Uses Hdlm

    • Changing the instance number of the HBA driver by changing the contents of the /etc/path_to_inst file • Switching the kernel mode Precautions Regarding Changes to the Configuration of an HDLM Operating Environment Keep the following in mind when changing the configuration of an HDLM operating environment.

  • Page 264

    A LUN was changed. ¢ The HBA model was changed. ¢ The HBA mounting slot was changed. ¢ The contents of the /etc/path_to_inst file were altered to change ¢ the instance number of the HBA driver. The FCode was installed on an HBA where the FCode had not been ¢...

  • Page 265: Switching The Kernel Mode, When The Path Configuration Is Changed In A Boot Disk Environment

    Obtain path information from the instance number of the HDLM driver. Execute the following command to display the driver configuration information, and then refer to the instance field of the HDLM driver obtained in step 1 (the driver name of the HDLM driver is dlmfdrv). Identify the path used to determine the HDLM logical device file name of the LU from the FilterPath property.

  • Page 266: Dynamic Reconfiguration (dr) For Solaris, Overview Of Reconfiguring The Hdlm Device, Reconfiguring The Hdlm Device

    Migrate to the local boot disk environment, following the procedure described in Migrating from a Boot Disk Environment to the Local Boot Disk Environment on page 3-96. Change the path configuration. Return to the HDLM boot disk environment, following the procedure described in Configuring a Boot Disk Environment on page 3-76.

  • Page 267: Notes On Reconfiguring The Hdlm Device

    • Reconfiguring an LU that is not managed by HDLM, as an HDLM-managed device Reconfigure an LU that has already been added to the host, but is not managed by HDLM, as an HDLM-managed device. The following operations are not supported in the dynamic reconfiguration of the HDLM device.

  • Page 268

    new LUs, execute the dlmsetconf utility, and then execute the dlmcfgmgr utility for managing the HDLM configuration. • Without restarting the system, when you add a new path to the LU that HDLM manages, the logical device file of the sd or ssd device corresponding to the path will not be created.

  • Page 269: Adding A New Lu (when Not Using Sun Cluster)

    • After you execute the dlmcfgmgr utility to add a new LU, when you execute the path status display command (the view -path operation), the HDevName of the added LU may be displayed as a hyphen (-). Adding a New LU (When Not Using Sun Cluster) To add a new LU: Set up the storage system and switch so that Solaris recognizes the LU.

  • Page 270

    Figure 4-1 Example of executing the format command Execute the dlmsetconf utility for creating the HDLM driver configuration definition file to update the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf). # /opt/DynamicLinkManager/bin/dlmsetconf Set up LUs not to be managed by HDLM. If you do not want HDLM to manage one of the newly recognized LUs, set up the LU so that it is excluded from HDLM management.

  • Page 271: Adding A New Lu By Restarting The Nodes (when Using Sun Cluster)

    # /opt/DynamicLinkManager/bin/dlmcfgmgr -a In environments not supporting dynamic reconfiguration of HDLM devices, restart the host, with the reconfiguration option specified. # reboot -- -r Check the status of the path for accessing the added LU. The following shows an example of executing the command. The shaded portion represents the path that accesses the added LU.

  • Page 272

    To add a new LU: Set up the storage system and switch so that Solaris recognizes the LU. Check the LU newly recognized by the node. You can check the LU by using the format command. Figure 4-2 Example of executing the format command on page 4-26 shows an example of executing the format command.

  • Page 273

    # /usr/cluster/lib/sc/scsi -c inkeys -d /dev/rdsk/ c3t50060E80033A9902d2s2 If the key is displayed, perform the steps beginning with step 5. Even when you release the setup for Quorum devices and disk device groups, since the key used for disk reservation will remain registered, you will need to restart the node in non-cluster mode and unregister the key.

  • Page 274

    The device mentioned in Devices Managed by HDLM on page 2-3 will be set as an HDLM-managed device. When the dlmsetconf utility is executed, the KAPL10242-I message will be displayed prompting you to confirm the deletion of the logical device files for the sd or ssd device.

  • Page 275

    This message is output when HDLM deletes a logical device file for an sd or ssd device managed by HDLM. Perform the procedure for updating the device ID information (steps 17 to 20) so that this message is not output next time the node starts. For details, refer to the Sun Cluster documentation.

  • Page 276

    Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r Notify the node of the most recent device ID information. Execute the following command: # /usr/cluster/bin/scdidadm -ui Update the Sun Cluster global device namespace.

  • Page 277: Adding A New Lu Via Dynamic Reconfiguration (when Using Sun Cluster)

    # /usr/cluster/bin/scgdevs When using, as a Quorum device, an HDLM device of an LU newly recognized for the first time by HDLM driver, set up the Quorum device on any node in the cluster. The following shows an example of executing the command: Check the device ID of the HDLM device.

  • Page 278

    procedure described in Adding a New LU By Restarting the Nodes (When Using Sun Cluster) on page 4-25. For details on the precautions for the dlmcfgmgr utility, see ¢ dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-15. To add a new LU: Set up the storage system and switch so that Solaris recognizes the LU.

  • Page 279

    Execute the following command on any node: # /usr/cluster/lib/sc/scsi -c inkeys -d the-name-of-the-logical- device-file-on-slice-2-of-the-LU-newly-recognized-by-the-node The following shows an example of executing the command: # /usr/cluster/lib/sc/scsi -c inkeys -d /dev/rdsk/ c3t50060E80033A9902d2s2 Add the new LU to HDLM by following the steps in Adding a New LU By Restarting the Nodes (When Using Sun Cluster) on page 4-25beginning...

  • Page 280

    If you do not want HDLM to manage one of the LUs newly recognized by the node, set up the LU so that it is excluded from HDLM management. For details, see Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page 3-66.

  • Page 281

    Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r Notify the node of the most recent device ID information. Execute the following command: # /usr/cluster/bin/scdidadm -ui Update the Sun Cluster global device namespace.

  • Page 282: Configuration Changes Such As Deleting A Logical Unit

    # /usr/cluster/bin/scgdevs When using, as a Quorum device, an HDLM device of an LU newly recognized for the first time by HDLM driver, set up the Quorum device on any node in the cluster. The following shows an example of executing the command: Check the device ID of the HDLM device.

  • Page 283: Changing The Configuration By Restarting The Host

    • Changing the host LUN • Changing, adding, or deleting a CHA port • Changing a target ID • Adding or replacing a Fibre Channel switch • Replacing, deleting, or adding an HBA To change an LU, we recommend that you add a new LU and then delete the existing one, rather than changing a LU in the storage system management program.

  • Page 284

    If VxVM uses the LU you are deleting, delete the volumes and VxVM disk groups. To have another LU take over from the LU you are deleting, replace the LU. If you are using SDS or SVM, perform the following as required: If you are using a shared diskset, unregister the HDLM devices you ¢...

  • Page 285

    # /opt/DynamicLinkManager/bin/dlmsetconf The KAPL10242-I message appears when you execute the dlmsetconf utility. Type y in response. Execute the following command to compare the HDLM driver unconfiguration definition file before and after execution of the dlmsetconf utility: # diff dlmfdrv.unconfbakn dlmfdrv.unconf Compare the old and new files to make sure that the reconfigured environment is recognized by HDLM.

  • Page 286

    > ### Device=ssd83 target=134012 lun=3 > ### Path=/pci@8,700000/SUNW,emlxs@1/fp@0,0/ ssd@w50060e80033a9939,3 Execution example after a path is added by the addition of a CHA port: Check that a path for the same CHA port has been added: 26a31,32 > ### Device=ssd119 target=134012 lun=2 >...

  • Page 287

    boot -r (at the ok prompt) ¢ Execute the dlnkmgr command's view operation specifying the -lu parameter, and check that the configuration has changed. If you are using Sun Cluster, remove the device ID information of devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C If you are using Sun Cluster, assign device IDs.

  • Page 288: Deleting An Lu Via Dynamic Reconfiguration

    for details, see To Use a Local Metadevice on page 3-157 Registering HDLM Devices on page 3-157. Recreate any volumes you deleted in step 8. For details, see To Use a Local Volume on page 3-161 Registering HDLM Devices on page 3-161.

  • Page 289

    Execute the dlmcfgmgr utility for managing the HDLM configuration to delete the LU, which is disconnected from the host, from the HDLM driver. # /opt/DynamicLinkManager/bin/dlmcfgmgr -a The HDLM device will be deleted without restarting the host. Execute the dlnkmgr command's view -path operation to check the path status after the configuration change.

  • Page 290: Changing An Hdlm-managed Device To A Non-hdlm-managed Device

    Changing an HDLM-managed Device to a Non-HDLM-Managed Device To change an HDLM-managed device to a non-HDLM-managed device: Check the LDEV information of the LU to be removed from HDLM management. Use the following command to check the LDEV information: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv PathID HDevName Device LDEV 000000 c7t0d0 sd46...

  • Page 291

    Figure 4-5 Example of executing the format command In this example, the shaded portion indicates the path information of the LU to be managed by HDLM. Edit the HDLM driver unconfiguration definition file (/opt/ DynamicLinkManager/config/dlmfdrv.unconf). Using a text editor, edit the definitions in the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/dlmfdrv.unconf) for the LU to be managed by HDLM.

  • Page 292

    Figure 4-6 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (before editing) Figure 4-7 Example of editing the /opt/DynamicLinkManager/ config/dlmfdrv.unconf file The shaded portion indicate the text to be edited. Note Do not edit anything other than adding the hash mark (#) to the beginning of the LDEV property line. Apply the contents that were edited in step 2 to the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf).

  • Page 293

    Specifying Whether a Logical Unit Is To Be Managed by HDLM (When Using Sun Cluster) Changing an HDLM-Managed Device to a Non-HDLM-Managed Device Note When you remove an HDLM-managed LU from management, perform ¢ the remove operations on all nodes in the cluster, and make the configuration the same on each node.

  • Page 294

    specifies that the LU indicated by that LDEV attribute is not to be managed by HDLM. Figure 4-8 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (before editing) on page 4-48 Figure 4-9 The /opt/ DynamicLinkManager/config/dlmfdrv.unconf file (after editing) on page 4-48 show examples of editing the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/dlmfdrv.unconf).

  • Page 295

    Figure 4-10 Result output by executing the dlmvxexclude utility Figure 4-11 Coding in the /etc/vx/vxvm.exclude file The shaded portion indicates the device not to be managed by HDLM and to be released from the VxVM disabled state. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-26.

  • Page 296

    Note The following message might be output to the console window or syslog while the node is running: Unable to open '/dev/rdsk/c5t50060E80033A9902d2s2' This message is output when an HDLM-managed device has been changed to a non-HDLM-managed device. Perform the procedure for updating the device ID information (steps 12 to 15) so that this message is not output next time the node starts.

  • Page 297

    example of executing the format command. The shaded portion represents an LU not managed by HDLM. Figure 4-12 Example of executing the format command Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C Allocate the device ID.

  • Page 298

    The following shows examples of executing the commands: Check the device ID of the device not to be managed by HDLM: # /usr/cluster/bin/scdidadm -L Set up the Quorum device: # /usr/cluster/bin/scconf -a -q globaldev=d4 Execute the following command to clear the install mode of a cluster when it consists of two nodes: # /usr/cluster/bin/scconf -c -q reset Perform registration for the disk device group, create the file system, or...

  • Page 299

    If setup is being performed for Quorum devices and disk device groups in the LU to be managed by HDLM, release the LU settings. When you unregister the SDS or SVM diskset, the diskset data will be deleted so first back up any necessary data. For details on how to release settings for Quorum devices and VxVM disk groups, see Remove Sun Cluster Settings on page 3-166...

  • Page 300

    Using a text editor, edit the definitions in the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/dlmfdrv.unconf) for the LU to be managed by HDLM. Based on the path information of the LU to be managed by HDLM checked in step 2 (shaded portion), find the correct LU in the HDLM driver unconfiguration definition file (/opt/ DynamicLinkManager/config/dlmfdrv.unconf), and add a hash mark (#) in front of the LDEV property line corresponding to the LU.

  • Page 301

    Specify settings as shown in Removing Devices from VxVM on a Controller Basis on page 3-142 Removing Devices From VxVM on a Path Basis on page 3-146 Removing Devices from VxVM on page 3-140. Stop all nodes in the cluster. If the key is displayed in step 4, and the node is running in non-cluster mode in step 6, execute the following command on all nodes: # shutdown -g0 -y -i0...

  • Page 302

    Permanent KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss Even when cluster software is used, it is not displayed in Support ¢ Cluster. Nevertheless, no problem exists with the functionality for cluster support. Use the execution results from the view operation to check that the programs are operating normally.

  • Page 303

    Execute the following command: # /usr/cluster/bin/scdidadm -ui Update the Sun Cluster global device namespace. Execute the following command: # /usr/cluster/bin/scgdevs When using an HDLM-managed device, as a Quorum device, set the Quorum device on any node in the cluster. The following shows examples of executing the commands: Check the device ID of the HDLM device: # /opt/DynamicLinkManager/bin/dlnkmgr view -path # /usr/cluster/bin/scdidadm -L...

  • Page 304

    Changing a Non-HDLM-Managed Device to an HDLM-Managed Device (For Dynamic Reconfiguration) Note Dynamic reconfiguration cannot be performed when using VxVM ¢ because the host must be restarted in order to use (in VxVM) an LU that was changed from a non-HDLM-managed device to an HDLM- managed device.

  • Page 305

    For details on how to release SDS or SVM disksets, see Remove SDS Settings on page 3-169 Remove SVM Settings on page 3-170 Preparations for HDLM Uninstallation on page 3-165. Make sure that registration is performed for the key used in disk reservation for the LU to be managed by HDLM.

  • Page 306

    Figure 4-18 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (after editing) The shaded portion indicates the text to be edited. Note Do not make any modifications other than adding or deleting the hash mark (#) in front of the LDEV property line corresponding to the LU. Apply the settings in the HDLM driver unconfiguration definition file (/ opt/DynamicLinkManager/config/dlmfdrv.unconf) to the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf):...

  • Page 307

    Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r Notify the node of the most recent device ID information. Execute the following command: # /usr/cluster/bin/scdidadm -ui Update the Sun Cluster global device namespace.

  • Page 308

    If you are using VxVM, see Linking VxVM and Sun Cluster on page 3-154. If you are using SDS, see To Use a Shared Diskset on page 3-158 Registering HDLM Devices on page 3-157. If you are using SVM, see To Use a Shared Diskset on page 3-161 Registering HDLM Devices on page 3-161.

  • Page 309: Troubleshooting

    Troubleshooting This chapter describes how to properly use HDLM error information, and then how to resolve the problem, if an error occurs in HDLM. The resolutions for path errors, HDLM program errors, and other types of errors are all described separately.

  • Page 310: Checking Error Information In Messages

    Information collected by using the DLMgetras utility for collecting HDLM error information Immediately after an error occurs, execute the DLMgetras utility to collect HDLM error information, because restarting the machine may delete existing error information already collected by the DLMgetras utility. For details on the DLMgetras utility and the error information it collects, see The DLMgetras Utility for Collecting HDLM Error Information on page...

  • Page 311

    The path name indicates a physical path. When you modify the system configuration or replace a piece of hardware, you should check the path names to identify the physical paths that will be affected by the change. A path name consists of the following four elements, separated by periods: Host port number (hexadecimal) ¢...

  • Page 312

    Figure 5-1 Error Location When the KAPL08022-E Message is Output Figure 5-2 Troubleshooting Procedure When a Path Error Occurs on page shows the troubleshooting procedure when the KAPL08022-E message is output. Figure 5-2 Troubleshooting Procedure When a Path Error Occurs Troubleshooting Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 313: Identifying The Error Path, Narrowing Down The Hardware That Might Have Caused The Error

    If an error occurs in a boot disk environment on a path connected to the boot disk, perform the procedure described in Actions to Take for a Path Error in a Boot Disk Environment on page 5-6. The following shows the procedure for using the dlnkmgr command to handle a path error.

  • Page 314: Actions To Take For A Path Error In A Boot Disk Environment

    operation, see online (Places Paths Online) on page 6-11. Execute the following command: # /opt/DynamicLinkManager/bin/dlnkmgr online Executing this command places all the offline paths online. If any path cannot be placed online due to an error, the KAPL01039-W message will appear. To ignore such paths and to continue processing, type y.

  • Page 315: Path Errors After Boot Processing Completes, What To Do For A Program Error

    When a path error occurs before the HDLM manager starts If boot processing is cancelled due to errors on all paths to the boot disk HDLM does not output the KAPL08022-E message on detection of a path error. Investigate the error and take appropriate action, referring to the following information: The message output to the console by Solaris ¢...

  • Page 316: Examining The Messages, Obtaining Program Information, What To Do For The Program Error

    Figure 5-3 Troubleshooting Procedure When a Program Error Occurs The following shows the procedure for handling a program error by using the dlnkmgr command. Examining the Messages Examine the messages that are output to the host syslog. If an error occurs in an HDLM program, a message other than KAPL08xxx is output to syslog.

  • Page 317: Contacting Your Hdlm Vendor Or Maintenance Company, What To Do For Other Errors

    If the KAPL01012-E message appears as a result of executing the command The following shows the KAPL01012-E message: KAPL01012-E Could not connect the HDLM manager. Operation name = view If this message appears, start the HDLM manager. For details on how to start the HDLM manager, see Starting the HDLM Manager on page 4-15.

  • Page 318

    5-10 Troubleshooting Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 319: Command Reference

    Command Reference This chapter describes the HDLM command (dlnkmgr) and its operations. □ Overview of the HDLM Command dlnkmgr □ clear (Returns the Path Statistics to the Initial Value) □ help (Displays the Operation Format) □ offline (Places Paths Offline) □...

  • Page 320: Overview Of The Hdlm Command Dlnkmgr

    Overview of the HDLM Command dlnkmgr This section describes how to specify the HDLM command dlnkmgr and its subcommands (called operations in HDLM). Command format Enter the command using the following format: dlnkmgr operation [parameter [parameter-value]] dlnkmgr The command name. operation The type of operation entered after dlnkmgr.

  • Page 321: To Set The Path Statistics To 0, To Display The Format Of The Clear Operation

    Execute the command as a user with root permissions. ¢ To specify a value that contains a space in its parameter, enclose the ¢ entire value in double quotation marks ("). clear (Returns the Path Statistics to the Initial Value) The dlnkmgr command's clear operation clears the statistics (I/O count and I/O errors) of all paths that are managed by the HDLM, and returns them to their initial value.

  • Page 322: To Display The Format Of The Clear Operation, Parameter

    To display the format of the clear operation -help Displays the format of the clear operation. Example # /opt/DynamicLinkManager/bin/dlnkmgr clear -help clear: Format dlnkmgr clear -pdst [-s] KAPL01001-I The HDLM command completed normally. Operation name = clear, completion time = yyyy/mm/dd hh:mm:ss help (Displays the Operation Format) The dlnkmgr command's help operation displays the list of operations available for the dlnkmgr command, or the format of individual operations.

  • Page 323

    Examples Example 1 The following example shows how to display all the operations available for the dlnkmgr command. # /opt/DynamicLinkManager/bin/dlnkmgr help dlnkmgr: Format dlnkmgr { clear | help | offline | online | set | view | monitor } KAPL01001-I The HDLM command completed normally. Operation name = help, completion time = yyyy/mm/dd hh:mm:ss Example 2 The following example shows how to display the formats of multiple...

  • Page 324: To Place Paths Offline, To Display The Format Of The Offline Operation

    The following example shows how to display the operations that can be specified for the help operation # /opt/DynamicLinkManager/bin/dlnkmgr help help help: Format dlnkmgr help { clear | offline | online | set | view | monitor } KAPL01001-I The HDLM command completed normally. Operation name = help, completion time = yyyy/mm/dd hh:mm:ss offline (Places Paths Offline) The dlnkmgr command's offline operation places online paths offline.

  • Page 325: Parameters, To Place Paths Offline

    Parameters To place paths offline -path Indicates that the target of the operation is a path managed by HDLM. This parameter is optional because offline is always used for paths, so it is assumed. Make sure that you specify the paths to be taken offline by using the - hba, -hbaid, -cha, -chaid, -pathid, or -device parameter.

  • Page 326

    # /opt/DynamicLinkManager/bin/dlnkmgr offline -hbaid 00001 KAPL01102-I All the paths which pass the specified HBA port will be changed to the Offline(C) status. Is this OK? [y/n]:y KAPL01103-I If you are sure that there would be no problem when all the paths which pass the specified HBA port are placed in the Offline(C) status, enter y.

  • Page 327

    # /opt/DynamicLinkManager/bin/dlnkmgr view -cha For details about how to execute the view operation to display the CHA port IDs, see To display CHA port information on page 6-60 Parameters on page 6-32. Example The following example shows how to place offline all paths connected to the CHA port whose CHA-port-ID is 00001.

  • Page 328: To Display The Format Of The Offline Operation

    The following shows an example of how to place offline all paths whose SCSI device name is sd52 while confirming command operation. # /opt/DynamicLinkManager/bin/dlnkmgr offline -path –device sd52 KAPL01052-I The currently selected paths will be changed to the Offline(C) status. Is this OK? [y/n]:y KAPL01053-I If you are sure that there would be no problem when the path is placed in the Offline(C) status, enter y.

  • Page 329: To Place Paths Online

    HBA_ID { 00000 - 99999 }(Decimal) CHA_ID { 00000 - 99999 }(Decimal) KAPL01001-I The HDLM command completed normally. Operation name = offline, completion time = yyyy/mm/dd hh:mm:ss Reference Executing the view operation together with standard UNIX commands enables you to filter the path information listed for a specific HBA port or CHA port.

  • Page 330: To Display The Format Of The Online Operation, Parameters, To Place Paths Online

    [-hba host-port-number.bus-number |-hbaid HBA-port-ID |-cha -pathid AutoPATH_ID |-chaid CHA-port-ID |-pathid AutoPATH_ID |-device SCSI-device-name] [-s] To display the format of the online operation /opt/DynamicLinkManager/bin/dlnkmgr online -help Parameters To place paths online -path Indicates that the target of the operation is a path managed by HDLM. This parameter is optional because online is always used for paths, so it is assumed.

  • Page 331

    be changed to the Online status. Is this OK? [y/n]:y KAPL01061-I 3 path(s) were successfully placed Online; 0 path(s) were not. Operation name = online -hbaid HBA-port-ID Use this parameter to place online, at one time, all paths that pass through a specific HBA port.

  • Page 332

    The following example shows how to place online the paths connected to the CHA port 0A. In this example, a path whose AutoPATH_ID is 000002 is connected to the target CHA port: # /opt/DynamicLinkManager/bin/dlnkmgr online -cha -pathid 000002 KAPL01057-I All the paths which pass the specified CHA port will be changed to the Online status.

  • Page 333

    # /opt/DynamicLinkManager/bin/dlnkmgr view -lu For details on how to execute the view operation and display the SCSI device name, see To display LU information on page 6-50 Parameters on page 6-32. Only one case-sensitive parameter value can be specified. Example The following shows an example of how to place online all paths whose SCSI device name is sd52 while confirming command operation.

  • Page 334: To Set Up The Hdlm Operating Environment

    dlnkmgr online [-path] [-pathid AutoPATH_ID] [-s] dlnkmgr online [-path] [-device SCSI_Device_Name] [-s] Valid value AutoPATH_ID { 000000 - 999999 }(Decimal) HBA_ID { 00000 - 99999 }(Decimal) CHA_ID { 00000 - 99999 }(Decimal) KAPL01001-I The HDLM command completed normally. Operation name = online, completion time = yyyy/mm/dd hh:mm:ss Reference Executing the view operation together with standard UNIX commands enables you to filter the path information listed for a specific HBA port or CHA...

  • Page 335: To Display The Format Of The Set Operation, Parameters

    {-lb {on [-lbtype {rr|exrr|lio|exlio|lbk|exlbk}]|off} |-ellv error-log-collection-level |-elfs error-log-file-size |-elfn number-of-error-log-files |-systflv trace-level |-systfs trace-file-size |-systfn number-of-trace-files |-pchk {on [-intvl checking-interval]|off} |-afb {on [-intvl checking-interval]|off} |-iem { on [-intvl error-monitoring-interval] [-iemnum number-of-times-error-is-to-occur] | off } |-lic |-audlog {on [-audlv audit-log-data-collection-level] [-category [[ss] [a] [ca]|all]]|off} |-audfac facility-value} [-s]...

  • Page 336

    Item name Default value Recommended value Number of trace files Path health on (Interval is 30 minutes.) on (Interval is 30 minutes.) checking The recommended checking interval depends on the operating environment. Automatic failback The recommended checking interval depends on the operating (1-minute checking interval) environment.

  • Page 337

    Table 6-3 Values of the Error Log Collection Level Value Description Collects no error log. Collects error information at the Error level or higher. Collects error information at the Warning level or higher. Collects error information at the Information level or higher. Collects error information at the Information level or higher (including maintenance information).

  • Page 338

    The trace files for which trace levels can be set are hdlmtrn.log (n indicates a file number from 1 to 64). Table 6-4 Trace Level Values on page 6-20 shows the values of the trace level. When an error occurs, we recommend that you set the value of the trace level to 1 or higher and re-collect log information.

  • Page 339

    on: Enabled off: Disabled For a standby host, or a host connected to the Thunder 9500V series, or Hitachi AMS/WMS series storage system, we recommend that you activate path health checking so that you can detect errors in paths where I/Os operations do not occur. When you specify on, specify the checking interval of path health checking by specifying the parameter immediately following on.

  • Page 340

    To prevent intermittent errors from deteriorating I/O performance, we recommend that you also enable intermittent error monitoring when the automatic failback function is enabled. Intermittent error monitoring is specifiable only when automatic failbacks are enabled. For details on the relationship between automatic failbacks and intermittent error monitoring, see Table 6-5 Relationship Between the Setting for the Automatic Failback Function and Intermittent Error...

  • Page 341

    -iem { on [-intvl error-monitoring-interval] [-iemnum number-of-times- error-is-to-occur] | off } Enables or disables intermittent error monitoring. on: Enabled off: Disabled Intermittent error monitoring can be set when automatic failback is set to When you use automatic failback, we recommend that you set intermittent error monitoring to on to prevent an intermittent error from reducing I/O performance.

  • Page 342

    During intermittent error monitoring, if changes are made in the settings of the intermittent error monitoring interval, the error count and the elapsed time measured since monitoring starts are set to 0, and monitoring starts by using the new settings. Outside the duration of intermittent error monitoring, if changes are made in the settings of the intermittent error monitoring interval, the new settings take effect after the next time automatic failback...

  • Page 343

    on page 6-25 shows the relationship between the settings for the automatic failback function and intermittent error monitoring and the executable operations for those functions. Table 6-5 Relationship Between the Setting for the Automatic Failback Function and Intermittent Error Monitoring and the Executable Operations Setting Executable operation Result of operation...

  • Page 344

    Setting Executable operation Result of operation Set AFB to on. AFB and IEM are enabled. Set AFB to off. The operations of AFB and IEM do not change. Set AFB to on. AFB is enabled. Set AFB to off. The operations of AFB and IEM do not change.

  • Page 345

    A message confirming that the license key has been registered is displayed, depending on the license key type described in the license key file. For a temporary license key or emergency license key, the expiration period is also displayed (KAPL01071-I, KAPL01072-I). Table 6-6 License Key Types on page 6-27 lists the license key types.

  • Page 346

    off: Audit Log data is not collected. -audlv audit-log-data-collection-level Specifies the severity level of audit log data to be collected. Table 6-7 Values Indicating Audit Log Data Collection Levels on page 6-28 describes the values that indicate audit log data collection levels. The default is 6.

  • Page 347

    Table 6-9 Values of the Audit Log Facility Value Corresponding facility value in the /etc/syslog.conf file user or 1 user local0 or 16 local0 local1 or 17 local1 local2 or 18 local2 local3 or 19 local3 local4 or 20 local4 local5 or 21 local5 local6 or 22...

  • Page 348: View (displays Information)

    [ -audlv AudlogLevel ] [ -category Category-Value ] | -audlog off | -audfac { Facility-Name | Facility-Number } [-s] Valid value ElogLevel { 0 | 1 | 2 | 3 | 4 } (Default Value 3) ElogFileSize { 100 - 2000000 }(KB) (Default Value 9900) Number-Of-ElogFiles { 2 - 16 }(Files)

  • Page 349: Format, To Display Program Information, To Display Path Information, To Display Lu Information

    Format To display program information /opt/DynamicLinkManager/bin/dlnkmgr view -sys [-sfunc|-msrv|-adrv|-pdrv|-lic|-audlog] [-t] To display path information To display path information /opt/DynamicLinkManager/bin/dlnkmgr view -path [-hdev host-device-name] [-stname] [-iem] [-srt {pn|lu|cp}] [-hbaportwwn] [-t] To display path information (by selecting a display item) /opt/DynamicLinkManager/bin/dlnkmgr view -path -item [pn] [dn] [lu] [cp] [type] [ic] [ie] [dnu] [hd] [iep] [hbaportwwn] [-hdev host-device-name]...

  • Page 350: To Display Hba Port Information, To Display Cha Port Information, Parameters

    To display LU information (by adding items to be displayed) /opt/DynamicLinkManager/bin/dlnkmgr view -lu -item [ [slpr] [pn] [cp] [clpr] [type] [ic] [ie] [dnu] [iep] [pd] | all ] [-hdev host-device-name | -pathid AutoPATH_ID] [-t] To display a summary of LU information /opt/DynamicLinkManager/bin/dlnkmgr view -lu -c [-t] To display a summary of LU information (by adding items to be displayed)

  • Page 351: To Display Program Information

    To display corresponding information about an HDLM device, sd or ssd device, and LDEV on page 6-61 To display view operation format on page 6-62 To display program information -sys [-sfunc|-msrv|-adrv|-pdrv|-lic|-audlog] Displays the HDLM program information. Use one of the sub-parameters (following -sys) to specify the program information that you want to display.

  • Page 352

    Parameter and program information Item Description to be displayed Support Blank Cluster Error log collection level: Elog Level • 0: Collects no error information. • 1: Collects error information at the Error level or higher. • 2: Collects error information at the Warning level or higher.

  • Page 353

    Parameter and program information Item Description to be displayed When the setting of the automatic failback is on, the checking interval of automatic failback is displayed within the parentheses following on. The time is in minutes. • Setting for intermittent error monitoring: Intermittent Error Monitor on: Enabled...

  • Page 354

    Parameter and program information Item Description to be displayed Information • Permanent about the • Temporary HDLM license • Emergency License expiration: Expiration • When using a permanent license: - • When using a temporary license or emergency license: The license expiration period is displayed in the following format: yyyy/mm/dd(ndays after).

  • Page 355

    When you use cluster software, the settings of the cluster support function and the kinds of cluster servers are not displayed. However, the cluster support function will operate normally. Examples Example 1 The following example shows how to display information about the HDLM function settings: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -sfunc HDLM Version...

  • Page 356: To Display Path Information

    HDLM Driver Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss Example 5 The following example shows how to display information about the HDLM license: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -lic License Type Expiration Permanent KAPL01001-I The HDLM command completed normally.

  • Page 357

    -hdev host-device-name Displays information only for the paths accessing the specified host device. Specify the name of the logical device file (minus the slice number, in cUtXdY format) for the HDLM device to indicate the desired host device. The host-device-name string is case-sensitive. -stname Use this parameter to display the model ID of the storage system in the product ID element of the DskName field.

  • Page 358

    To display path information (by selecting a display item) -path -item When you specify the -path parameter together with the -item parameter, the command only displays the items specified by the value of the -item parameter. If you specify no value for the -item parameter, only the PathID and the Status fields are displayed.

  • Page 359

    Selectable items Sub-parameters Because both PathID and Status are always displayed, you do not have to specify any parameters. In the sub-parameters following -path -item, you can list the paths (- hdev) and sort the list (-srt). If you omit both parameters, the command displays information for all the paths in ascending order of AutoPATH_IDs.

  • Page 360

    When you use this parameter, the items used for the sorting keys (DskName, AutoPATH_ID, and the item specified by this parameter) are displayed by default. Therefore, it is not necessary to specify these items for the -item parameter. Omits the title for each information item. Example In the following example, IO-Count is selected as the display item and the path information is sorted in ascending order by LUs.

  • Page 361

    When you use the -c parameter, the number of characters that can be displayed in the product ID element of the DskName field is limited to 10. When there are 11 or more characters in the product ID, the 8th and following characters are abbreviated to an ellipsis (...).

  • Page 362

    Displayed path information Table 6-12 Displayed Path Information on page 6-44 describes the displayed path information. The following explains the table headings: • No summary displayed: The user specifies only the -path parameter or the -path -item parameter. • Summary displayed: The user specifies the -path -c parameter. Table 6-12 Displayed Path Information Displayed Item Description...

  • Page 363

    Displayed Item Description Summary summary displayed displayed Total I/O count for the displayed paths, indicated by a IO-Count decimal number. The maximum value that can be displayed is 2 - 1 (4294967295). If the total I/O count reaches the maximum value, it will re-start from 0. Total I/O error count for the displayed paths, indicated by a IO-Errors decimal number.

  • Page 364

    Displayed Item Description Summary summary displayed displayed This number combined with the storage system name (shown in DskName) identifies the LU that is accessed by a path. • For the Lightning 9900 Series, Lightning 9900V Series, Hitachi NSC55, Hitachi Universal Storage Platform 100, Hitachi Universal Storage Platform 600, or Hitachi Universal Storage Platform 1100, indicated by a hexadecimal number.

  • Page 365

    Displayed Item Description Summary summary displayed displayed When connecting to Lightning 9900 Series, Lightning 9900V Series, Hitachi USP, Universal Storage Platform V/VM, Virtual Storage Platform, Hitachi AMS2000 series, or Hitachi SMS all paths are owner paths. Total I/O count for the path, indicated by a decimal IO-Count number.

  • Page 366

    Displayed Item Description Summary summary displayed displayed Indicates that an intermittent error occurred (automatic failbacks do not check the path) (the path status is Online, Onlin(E), Offline(E), or Offline(C)). A 16-digit hexadecimal number indicating the WWN HBAPortWW information for an HBA connected to the storage system. This item is displayed only when you specify -iem together with the -hbaportwwn parameter.

  • Page 367

    Product ID With the -stname Model names of storage Without the -stname parameter systems parameter (Displays the following for the model ID) Hitachi AMS2000/AMS Series Hitachi WMS Hitachi SMS Lightning 9900 Series 9910/9960 Emulation type Lightning 9900V Series 9970/9980 • Hitachi NSC55 •...

  • Page 368: To Display Lu Information

    To display LU information When displaying LU information, if the -item parameter, -c parameter, or the -c -item parameter is specified at the same time as the -lu parameter, you can add and display items and display a summary of LU information. This section describes each parameter and the LU information and displayed items.

  • Page 369

    Product : 9970/9980 SerialNumber : 15001 HDevName Device PathID Status 0670 c6t5d0 sd74 000000 Online sd80 000006 Offline(E) 0671 c6t5d1 sd75 000001 Online sd81 000007 Offline(E) 0672 c6t5d2 sd76 000002 Online sd82 000008 Offline(C) KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss To display LU information (by adding items to be displayed) -lu -item...

  • Page 370

    In the sub-parameters, -hdev or -pathid, you can filter the LU information to be displayed. If you omit both parameters, the command displays the information about all the LUs recognized by HDLM. For details on the contents of each displayed item, see Table 6-17 Displayed LU Information on page 6-55.

  • Page 371

    To display a summary of LU information -lu -c When the -c parameter is specified with the -lu parameter, a summary of LU configuration information is displayed on one line. The total number of paths recognized by HDLM and the number of online paths are displayed for each LU.

  • Page 372

    0672 c6t5d2 KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss To display a summary of LU information (by adding items to be displayed) -lu -c -item The items specified with the -item parameter are added to the items to be displayed by the -lu -c parameter and displayed.

  • Page 373

    • Summary displayed: The user specifies the -lu -c parameter or -lu -c - item parameter. Table 6-17 Displayed LU Information Displayed item Description No summary Summary displayed displayed Model ID of the storage system Product Serial number of the storage system Serial number Total number of LUs managed by HDLM among the...

  • Page 374

    Displayed item Description No summary Summary displayed displayed The name of the logical device file (minus the slice number, in cUtXdY format) for the HDLM device U: Controller number reserved by HDLM. X: Target ID or WWN (World Wide Name) of the sd or ssd device for the HDLM device Y: LUN of the sd or ssd device for the HDLM device Type of SCSI device (sd or ssd) and corresponding...

  • Page 375

    Displayed item Description No summary Summary displayed displayed • Offline(E): Offline due to an error • Online(E): Failure has occurred (If none of the paths accessing one LU have an Online status, one of those paths is changed to the Online(E) status.) Paths that are Offline(E) or Online(E) require corrective action.

  • Page 376

    Displayed item Description No summary Summary displayed displayed • -: Indicates that intermittent error monitoring is disabled or the monitoring time for an intermittent error is out of range. • A value of at least 0: Indicates the number of errors that occurred during intermittent error monitoring.

  • Page 377

    Items Displayed as HBA Port Information on page 6-59 lists and describes the displayed items. Table 6-18 Items Displayed as HBA Port Information Item Description HBA port ID HbaID Port (host port number in hexadecimal) and Bus (bus number in Port.Bus hexadecimal), separated by a period A 16-digit hexadecimal number indicating the WWN information...

  • Page 378: To Display Cha Port Information

    Does not display the SCSI device configuration information (Physical Device). -portwwn Displays port WWN information for HBAs connected to the storage system using a 16-digit hexadecimal number. Omits the title for each information item. To display CHA port information -cha Use this parameter to display information about the paths that pass through each CHA port.

  • Page 379

    The following shows the subsequent sub-parameters: -srt cp Sorts the CHA port information by the storage system model IDs (Product), storage system serial numbers (S/N), and CHA port numbers (ChaPort). Omits the title for each information item. To display corresponding information about an HDLM device, sd or ssd device, and LDEV -drv Displays the information about the PathID, the HDLM device, the sd or...

  • Page 380

    Item Description Type of SCSI device (sd or ssd) and corresponding Device instance number The model ID, serial number, and iLU number for the LDEV storage system, separated by periods. Information about LDEV can identify an actual LU that is managed by HDLM. Example # /opt/DynamicLinkManager/bin/dlnkmgr view -drv PathID HDevName Device LDEV...

  • Page 381: To Display I/o Information For Each Hba Port

    [ic] [ie] [dnu] [iep] [pd] | all ] [ -hdev HostDeviceName | -pathid AutoPATH_ID ] [-t] dlnkmgr view -lu -c [-t] dlnkmgr view -lu -c -item [slpr] [-t] dlnkmgr view -drv [-t] dlnkmgr view -hba [-srt pb] [-c | -portwwn] [-t] dlnkmgr view -cha [-srt cp] [-t] KAPL01001-I The HDLM command completed normally.

  • Page 382

    Parameters Note If the number of paths to be displayed by the monitor operation is 500 or more, the execution of the monitor operation might be delayed as follows: The time displayed in StartTime might be later than the time the ¢...

  • Page 383

    Specifies the interval at which I/O information is to be displayed, in the range from 1 to 60 (seconds). When this parameter is omitted, the default (3 seconds) is applied. -count count Specifies the number of times I/O information is to be displayed, in the range from 1 to 999.

  • Page 384: To Display I/o Information For Each Cha Port

    To display I/O information for each CHA port -chaid CHA-port-ID Specifies the CHA port ID of the CHA port for which I/O information is to be displayed. The subsequent parameters specify the display interval and count. The format for the subsequent parameters are as follows: -intvl interval Specifies the interval at which I/O information is to be displayed, in the range from 1 to 60 (seconds).

  • Page 385: To Display Monitor Operation Format

    SerialNumber : 1100 ChaPort : 1R StartTime : yyyy/mm/dd hh:mm:ss ElapsedTime IO-Count IO-Errors 00:00:30 1154 00:01:00 00:01:30 KAPL01001-I The HDLM command completed normally. Operation name = monitor, completion time = yyyy/mm/dd hh:mm:ss To display monitor operation format -help Displays the monitor operation format. Example # /opt/DynamicLinkManager/bin/dlnkmgr monitor -help monitor:...

  • Page 386

    6-68 Command Reference Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 387: Overview Of The Utilities

    Utility Reference This chapter explains the utilities used by HDLM. □ Overview of the Utilities □ The DLMgetras Utility for Collecting HDLM Error Information □ The dlmcfgmgr Utility for Managing the HDLM Configuration □ The dlminstcomp HDLM Component Installation Utility □...

  • Page 388: Overview Of The Utilities

    Overview of the Utilities HDLM provides the following utilities: • The DLMgetras utility for collecting HDLM error information When an error occurs, this utility collects the files that contain information to be submitted to your HDLM vendor or maintenance company. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page...

  • Page 389: The Dlmgetras Utility For Collecting Hdlm Error Information

    dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-26. • The installhdlm utility for installing HDLM This utility performs a new installation, upgrade installation, or re- installation of HDLM in unattended installation mode. For details on this utility, see The installhdlm Utility for Installing HDLM on page 7-28.

  • Page 390

    [-f name-of-a-file-for-defining-the-information-to-be-collected]|-h} You can also use lower-case characters (dlmgetras) as follows: /opt/DynamicLinkManager/bin/dlmgetras {directory-to-which-collected- information-is-output [-f name-of-a-file-for-defining-the-information-to-be-collected]|-h} Parameters directory-to-which-collected-information-is-output Specify the output directory for the information that is to be collected by the DLMgetras utility for collecting HDLM error information. The collected information is compiled into three files shown in Function and output in the specified directory.

  • Page 391: List Of Collected Error Information, Format, Parameters

    Only one file or directory can be specified per line. ¢ The root directory (/) cannot be specified. ¢ When a directory is specified, the DLMgetras utility collects all the files ¢ in that directory, including files contained in the directory's subdirectories.

  • Page 392

    This file contains compressed error information of the Hitachi Command Suite products other than HDLM. This file is output only when using the Hitachi Command Suite product that is the target for collecting error information. • getras.tar.Z This file contains HDLM information and system information. When you execute the DLMgetras utility, specifying the file for defining information to be collected, the getras.tar.Z file contains the information defined in the file for defining information to be collected.

  • Page 393

    Files Explanation Output directory Previous generation of the HDLM dlmfdrv.unconf.ba driver unconfiguration definition file. dlmfdrv.unconf.bak is created if the dlmfdrv.unconf file already exists when you attempt to create the dlmfdrv.unconf file. Backup of the existing dlmfdrv.unconf_ol dlmfdrv.unconf file, taken when dbackup you perform an upgrade installation from HDLM 5.6.1 or...

  • Page 394

    Files Explanation Output directory you perform an upgrade installation from HDLM 5.6.1 or earlier. Backup of the existing dlmfdrv.conf.bak_ dlmfdrv.conf.bak file, taken oldbackup when you perform an upgrade installation from HDLM 5.6.1 or earlier. HDLM nexus driver configuration dlmndrv.conf definition file Backup file of the HDLM nexus dlmndrv.conf.bak driver configuration definition file...

  • Page 395

    Files Explanation Output directory OS crash dump configuration dumpadm.conf definition file I/O fencing function setup files gabtab llthosts llttab vxfendg vxfenmode vxfentab File system mount information mnttab syslog output directory definition syslog.conf Kernel patch file system Startup-time file system auto- vfstab mount information Solaris settings and configuration...

  • Page 396

    Files Explanation Output directory dlmgetomtrace.dmp In-memory trace information List of directories and files related dlmls-la.txt to the HDLM installation file information (output results from the ls -alR command) Output results of dlnkmgr view - dlmmgr-cha.txt Output results of dlnkmgr view - dlmmgr-hba.txt hba -portwwn HDLM LU information...

  • Page 397

    Files Explanation Output directory Process information ps-ef.txt CPU information psrinfo-v.txt List of installed patches showrev-p.txt Output result of the svcs -a svcs-a.txt command (Solaris 10) Output result of svcs -l svcs-l.txt hdlmload, svcs -l hdlm-manager, svcs -l hdlm-initpath and svcs -l hdlm-onlinepath (Solaris 10) Swap region information swap-l.txt...

  • Page 398

    Files Explanation Output directory Output results of vxdisk path VXVM_vxdiskpath.t VXVM_vxdglist.txt List of disk groups vxprint output results VXVM_vxprint.txt All directories and VxVM configuration information etc/vx files in the /etc/vx directory Solaris configuration information var/sadm/system/admin CLUSTER Log of the hdlmload service /var/svc/log system-device- hdlmload:default.

  • Page 399

    Files Explanation Output directory I/O fencing GAB information IOF_gabconfig- a.txt I/O fencing LLT link details IOF_lltstat- nvv.txt IOF_lltstat-p.txt I/O fencing LLT port information I/O fencing package information IOF_pkginfo.txt I/O fencing settings IOF_vxfenadm- d.txt VCS package information VCS_pkginfo.txt VCS status information VCS_hastatus- summary.txt VCS cluster information...

  • Page 400

    Files Explanation Output directory Status database information SVM_metadb.txt Volume information SVM_metastat.txt Diskset information SVM_metaset.txt metastat -s diskset-name SVM_metastat- s.txt information SDS input parameter file md.tab etc/opt/SUNWmd SDS configuration backup file md.cf Information on SDS status mddb.cf database copy storage locations (Not applicable) Sun Cluster information collection SC3.0...

  • Page 401

    This setup file is for VCS version 5.0. This information is collected only if the file exists. This information is collected only if the software for which information is to be collected is installed. The dlmcfgmgr Utility for Managing the HDLM Configuration This utility makes the HDLM driver recognize the contents of the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf) dynamically without restarting the host.

  • Page 402

    The dlmcfgmgr utility is valid in Solaris 8 version 1/01 or later, Solaris 9, or Solaris 10 system. Format /opt/DynamicLinkManager/bin/dlmcfgmgr { -a | -h } Parameters Applies the contents of the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf) to HDLM driver. The added LU becomes valid after the command with this option is executed.

  • Page 403

    The correspondence between the paths and path IDs might differ ¢ depending on whether dynamic reconfiguration of the HDLM device is performed by executing the dlmcfgmgr utility or by restarting the host. You must restart the host after performing an upgrade installation or ¢...

  • Page 404

    In Solaris 8 or 9, if the HDLM driver configuration definition file (/ ¢ kernel/drv/dlmfdrv.conf) does not contain HDLM device information, the dlmcfgmgr utility outputs the message and terminates. In Solaris 10, all LUs are deleted from the HDLM driver. If no change is made to the HDLM-managed devices when the ¢...

  • Page 405

    Format /opt/DynamicLinkManager/bin/dlmlisthdev input-file-name Parameters input-file-name Specifies, as an input file, the name of the output file that stores the output result of the dlnkmgr command's view -path operation executed before HDLM 04-00 or 04-01 was uninstalled. You cannot specify multi- byte characters.

  • Page 406

    the result in the standard output. The result of the command execution is as follows: Table 7-3 Information Displayed by the dlmlisthdev Utility for Assisting HDLM Transitions on page 7-20 lists and describes the items to be displayed. Table 7-3 Information Displayed by the dlmlisthdev Utility for Assisting HDLM Transitions Item Explanation...

  • Page 407: Configuration Definition File

    The dlmsetboot Utility for Assisting Configuration of an HDLM Boot Disk Environment This utility automates part of the tasks for configuring a boot disk environment in a Solaris 9 or later environment. Note that this utility cannot be used in an environment in which cluster software and a volume manager are used.

  • Page 408

    configuration definition file (/kernel/drv/dlmfdrv.conf) and the HDLM nexus driver configuration definition file (/kernel/drv/dlmndrv.conf). The sd or ssd device entries are sorted in ascending order of the LDEV information (for example, 9970/9980.15001.066A). If an existing HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf) is available, the existing definitions are inherited.

  • Page 409

    Parameters -d output-directory-name Specifies the name of the directory to which the configuration definition file created by the dlmsetconf utility is to be output. This directory must be specified as an absolute path. You cannot specify multi-byte characters. If output-directory-name is omitted, the configuration definition file will be created in the /kernel/drv directory.

  • Page 410

    dlmfdrv.unconf) contains an invalid entry or does not exist, an error occurs. This parameter cannot be specified together with the -r parameter. Examples Example 1 Parameters are omitted: A configuration definition file is created in the /kernel/drv directory. If the configuration definition file already exists, the existing definition information is inherited, and the parts for the changed configuration are added and/or deleted.

  • Page 411

    HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/ config/dlmfdrv.unconf), and the HDLM nexus driver configuration definition file (/kernel/drv/dlmndrv.conf). Entering n in response to this message does not create or update those files. • When the reconfiguration option is specified to restart the host after executing the dlmsetconf utility, HDLM deletes the HDLM-managed logical device files for sd or ssd devices.

  • Page 412: Configuration File

    the dlmsetconf utility is executed before the logical device files of the HDLM devices corresponding to the deleted LU are deleted, then the format command might incorrectly display the logical device file name of an HDLM device corresponding to the deleted LU, for the newly recognized LU.

  • Page 413

    creation of the VxVM configuration file outputs the contents of the file to the standard output. Format /opt/DynamicLinkManager/bin/dlmvxexclude [-d output-directory-name| -h ] Parameters -d output-directory-name Specifies the path name of the directory to which the file is to be output. You cannot specify multi-byte characters.

  • Page 414

    Only one HDLM device is created for one HDLM-managed device (one slice). However, VxVM assumes that the number of HDLM devices created for one HDLM-managed device is the same as the number of physical paths. Thus, executing the dlmvxexclude utility outputs entries for HDLM devices that have not actually been created.

  • Page 415: Contents Of The Installation-information Settings File

    on page 3-49, Performing a New Installation of HDLM (When Sun Cluster Is Being Used) on page 3-56, or Performing an Upgrade Installation or Re- installation of HDLM on page 3-67. Format /directory-where-DVD-ROM-is-mounted/hdlmtool/instutil/installhdlm { -f installation-information-settings-file-name | -h} Parameters -f installation-settings-file-name Defines the information required to perform an installation.

  • Page 416: About The Log File

    Required or optional? Maximu m length upgrade Key name Explanation For new (characte installatio installati n or re- installatio The default is /cdrom/ cdrom0. Specify the absolute path Optional Optional logdir name of the directory to which log data will be output.

  • Page 417

    Required or optional? Maximu m length upgrade Key name Explanation For new (characte installatio installati n or re- installatio n: Do not restart the system (default). If you use Sun Cluster, specify n. Legend: Optional: If you omit the key name and value, the default is used. However, for an upgrade installation or re-installation, the previous license information will be inherited for the licensekeyfile key.

  • Page 418

    Table 7-5 Keys that Can Be Defined in the [ENVIRONMENT_SETTINGS] Section Required or optional? Required Maximum length Required Key name Explanation upgrade (characte for new installati installatio on or re- installati Specify whether to enable Optional Optional load_balance load balancing by specifying either of the following values: on: Enable load balancing (default).

  • Page 419

    Required or optional? Required Maximum length Required Key name Explanation upgrade (characte for new installati installatio on or re- installati You can specify a value from 2 to 16. The default is 2. Specify the output level for Optional Optional trace_level traces.

  • Page 420

    Required or optional? Required Maximum length Required Key name Explanation upgrade (characte for new installati installatio on or re- installati Specify the path status check Optional Optional auto_failback_int interval, which is the interval erval between the end of the previous check and the start of the next check, in minutes.

  • Page 421: About The Log File

    Specifying an invalid value causes an error. For details on the settings functionality, see set (Sets Up the Operating Environment) on page 6-16. You can specify this key only in the following cases: For a new installation: ¢ When on is specified for the auto_failback key in the installation information settings file For an upgrade installation or re-installation: ¢...

  • Page 422: Log File

    • The installhdlm.log file is created in the directory specified by the logdir key in the installation-information settings file. Note that the directory must not be shared with hosts on which unattended installation is performed. If the logdir key does not have a value, the file is created in the /var/tmp directory.

  • Page 423

    file is created and named hdlm_installux_sh.log, to which the execution information is output. Note The hdlm_installux_sh.log and hdlm_installux_sh2.log log files are not deleted when HDLM is uninstalled. Manually delete the log files, if they are no longer required. 7-37 Utility Reference Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 424

    7-38 Utility Reference Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 425: Messages

    Messages This chapter describes the format and meaning of the message IDs, and also the terms used in the messages and message explanations. For details on the meaning of the return codes output by HDLM when it receives a request from Global Link Manager and measures to take for them, see Return Codes for Hitachi Command Suite Common Agent Component on page...

  • Page 426

    □ Return Codes for Hitachi Command Suite Common Agent Component Messages Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 427: Before Viewing The List Of Messages, Format And Meaning Of Message Ids

    Before Viewing the List of Messages This section explains the following information that is needed to locate messages and understand the explanations in the sections from KAPL01001 to KAPL02000 on page 8-4KAPL01001 to KAPL02000. • Format and meaning of the message IDs •...

  • Page 428: Kapl01001 To Kapl02000, Kapl01001 To Kapl

    • HDLM management target KAPL01001 to KAPL02000 Message Message Text Explanation KAPL01001 The HDLM command completed Details normally. Operation name = The HDLM command completed aa...aa, completion time = bb...bb successfully. aa...aa: clear, help, monitor, offline, online, set, or view bb...bb: Year/month/day hour:minute:second Action...

  • Page 429

    Message Message Text Explanation operation, see help (Displays the Operation Format) on page 6-4. KAPL01006 A necessary parameter is not Details specified. Operation name = The specified operation does not aa...aa contain the necessary parameter. aa...aa: clear, set, monitor, offline, or view Action Execute help operation-name of the HDLM command (dlnkmgr) to...

  • Page 430

    Message Message Text Explanation aa...aa: view Action Execute the view operation of the HDLM command (dlnkmgr) to check whether the HDLM manager has started. Start the HDLM manager if it has not started, and then retry the HDLM command. For details on the view operation, see view (Displays Information) on page...

  • Page 431

    Message Message Text Explanation Execute the view operation of the HDLM command (dlnkmgr view -path) and check the value displayed in PathName. Use the two leftmost digits of PathName for the relevant HBA port, and then retry. For details on the view operation, seeview (Displays Information) on page...

  • Page 432

    Message Message Text Explanation Execute the view operation of the HDLM command (dlnkmgr view -path) to check the value displayed in HDevName. Specify a host device for the value of HDevName, and then retry. For details on the view operation, seeview (Displays Information) on page 6-30.

  • Page 433

    Message Message Text Explanation KAPL01021 Cannot execute the HDLM Details command due to insufficient Memory required for HDLM memory. command processing could not be allocated. Action Terminate unnecessary applications to increase the amount of free memory, and then retry. KAPL01023 The last Online path for the device Details cannot be placed Offline(C).

  • Page 434

    Message Message Text Explanation continue the processing of the To ignore this path and perform online operation? [y/n]: online processing for the next path, enter y. To cancel processing, enter n. aa...aa: Path ID (decimal (base-10) number) Action If you want to continue processing of the online operation of the HDLM command for other paths, enter y.

  • Page 435

    Message Message Text Explanation Execute help operation-name of the HDLM command (dlnkmgr) to check the parameter value, and then retry. For details on the help operation, see help (Displays the Operation Format) on page 6-4). KAPL01048 Help information cannot be found. Details Operation name = aa...aa.

  • Page 436

    Message Message Text Explanation KAPL01051 Because no path has been selected Details among the currently displayed All the paths will be placed paths, the paths in the Offline(C), Online because the path Offline(E), and Online(E) statuses selection parameter is not will be changed to the Online specified for the online status.

  • Page 437

    Message Message Text Explanation a path that cannot be placed in the A path that cannot be placed set Offline(C) status was detected. to Offline(C) was detected during PathID = aa...aa Would you like to multipath offline processing. To continue the processing of the ignore this path and perform offline operation? [y/n]: offline processing for the next...

  • Page 438

    Message Message Text Explanation the Online status. Is this OK? [y/ or -cha parameter was specified. If you want to continue, enter y. To cancel the operation, enter n. aa...aa: CHA port, HBA Action If you want to execute online processing for the paths that meet the specified requirements, enter y.

  • Page 439

    Message Message Text Explanation aa...aa: online, offline, set, or clear Action None. KAPL01061 aa...aa path(s) were successfully Details placed bb...bb; cc...cc path(s) were This message indicates the not. Operation name = dd...dd number of the paths processed in an online/offline operation. aa...aa: Number of paths where the online/offline operation was successful (decimal...

  • Page 440

    Message Message Text Explanation None. KAPL01069 The entered license key is invalid. Details The entered license key is invalid. Action Enter a valid license key. KAPL01070 The entered license key is invalid. Details Renewal of the license key will now The license key renewal stop.

  • Page 441

    Message Message Text Explanation Re-install HDLM. KAPL01076 The permanent license has been Details installed. You need not install a license because a permanent license has already been installed. Action None. KAPL01078 The operation terminated because Details the path configuration changed aa...aa: monitor during execution of the HDLM Action...

  • Page 442

    Message Message Text Explanation KAPL01082 There is no installable license key Details in the license key file. File name = There is no useable license key aa...aa for HDLM in the license key file. aa...aa: /var/tmp/hdlm_license Action Make sure that the license key file is correct, and then re- execute.

  • Page 443

    Message Message Text Explanation KAPL01089 One of the following was executed Action at the same time as an HDLM Check the license by using the command set -lic operation: HDLM command's view -sys - another set -lic operation, or an lic operation.

  • Page 444

    Message Message Text Explanation to be put online has been reserved by another path. aa...aa: Path ID bb...bb: Code showing reason for error Action Use the format command to confirm that the LU for the path to be put online is not reserved by another host.

  • Page 445

    Message Message Text Explanation HDLM error information to collect the error information, and then contact your HDLM vendor or maintenance company if you have a maintenance contract. KAPL01095 An attempt to acquire the HDLM Details version information has failed. aa...aa: Code showing the reason details = aa...aa for the error Action...

  • Page 446

    Message Message Text Explanation Action None. KAPL01101 The target HBA port was not found. Details Operation name = aa...aa The HBA having the HBA_ID specified in the -hbaid parameter could not be found. aa...aa: monitor, offline, online Action Use the view operation of the HDLM command (dlnkmgr view -hba) to check the target HBA port and the HBA_ID of the target...

  • Page 447

    Message Message Text Explanation or -chaid parameter was specified. To collectively place multiple paths Online, enter y. To not collectively place them Online, enter n. aa...aa: CHA, HBA Action If you want to execute the Online processing for the paths which pass the specified target, enter y.

  • Page 448

    Message Message Text Explanation KAPL01115 The audit log configuration file Details cannot be opened. Execute the The audit log configuration file "dlnkmgr view -sys -audlog" cannot be opened. command and check whether a Action normal result is displayed. If the "dlnkmgr view -sys - audlog"...

  • Page 449: Kapl03001 To Kapl04000, Kapl03001 To Kapl

    Message Message Text Explanation KAPL01120 A storage system model ID could Details not be displayed. Details = aa...aa, A storage system model ID could bb...bb not be displayed. aa...aa: Storage recognition information bb...bb: Error code Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the...

  • Page 450

    Message Message Text Explanation Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see DLMgetras Utility for Collecting HDLM Error Information on page 7-3.

  • Page 451

    Message Message Text Explanation Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see DLMgetras Utility for Collecting HDLM Error Information on page 7-3.

  • Page 452: Kapl04001 To Kapl05000, Kapl04001 To Kapl

    KAPL04001 to KAPL05000 Message Message Text Explanation KAPL04001 HDLM manager started. Action None. KAPL04002 Could not start the HDLM manager. Details The HDLM manager failed to start because the current environment is unsuitable for the HDLM manager to run in. Action Execute the DLMgetras utility for collecting HDLM error...

  • Page 453

    Message Message Text Explanation file has been inadvertently deleted. KAPL04009 The option definition is invalid. Details HDLM manager could not start normally because some of the definitions in the option definition file were invalid. Action If the KAPL04033-W message is output after this message, execute the dlnkmgr view -sys -sfunc command, and then...

  • Page 454

    Message Message Text Explanation KAPL04012 Could not create a communication Details pipe. RC = aa...aa HDLM manager could not start normally because it was unable to create a pipe file, which is used in communication with HDLM commands. aa...aa: OS error code (decimal (base-10) number) Action Execute the DLMgetras utility for...

  • Page 455

    Message Message Text Explanation a maintenance contract for HDLM. For details on the DLMgetras utility, see DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL04019 Could not collect the error Details information. RC = aa...aa An attempt to read the log information from the alert driver failed.

  • Page 456

    Message Message Text Explanation KAPL04023 HDLM manager error information - Details aa...aa This information is required for resolving problems. aa...aa: HDLM manager trace error information Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.

  • Page 457

    Message Message Text Explanation Install a permanent license by the expiration day. KAPL04027 The emergency license is valid. The Details license expires in aa...aa days on aa...aa: Number of days (bb...bb). remaining bb...bb: Expiration date format: Year (4 digits)/month (01-12)/ day (01-31) Action Install a permanent license by...

  • Page 458

    Message Message Text Explanation DynamicLinkManager/ dlmmgr.xml) using the default values has failed. Action Remove unnecessary files to secure free capacity on the file system, or check the write permissions for the directory and file. KAPL04035 The path health check will now Details start.

  • Page 459: Kapl05001 To Kapl06000, Kapl05001 To Kapl

    Message Message Text Explanation None. KAPL05001 to KAPL06000 Message Message Text Explanation KAPL05003 The HDLM driver (filter component) Details was successfully attached to Disk The path for the disk and (aa...aa), Partition (bb...bb). partition was registered successfully in the core logic. aa...aa: minor number of the sd or ssd device (decimal number) bb...bb: slice number...

  • Page 460

    Message Message Text Explanation cc...cc: Error code (hexadecimal number) dd...dd: Address of table for managing the HDLM filter driver (hexadecimal number) Action Check whether the HDLM driver has started normally. If it has not started or contains an error, contact your HDLM vendor or the maintenance company, if there is a maintenance contract for HDLM, and report the error and...

  • Page 461

    Message Message Text Explanation bb...bb: Path ID (hexadecimal number) Action None. KAPL05020 Processing of IOCTL(aa...aa) will Details now start. An IOCTL request was received.. aa...aa: IOCTL code (hexadecimal number) Action None. KAPL05021 Processing of IOCTL(aa...aa) Details completed normally. The processing for the requested IOCTL operation was successful.

  • Page 462

    Message Message Text Explanation path is correctly connected to the aa...aa: Error code (decimal disk. number) bb...bb: IOCTL code (hexadecimal number) cc...cc: Minor number of the sd or ssd driver (hexadecimal number) Action Check whether the path is correctly connected to the disk. KAPL05040 The disk geometry or partition Details...

  • Page 463

    Message Message Text Explanation KAPL05042 Results of Persistent Group Details Reservation key registration - minor This message is generated for = aa...aa, ioctl = bb...bb, errno = maintenance. cc...cc, status = dd...dd This message shows the results of Persistent Group Reservation key registration or unregistration.

  • Page 464

    Message Message Text Explanation Refer to the HDLM User's Guide for aa...aa: instance number instructions how to execute the (decimal number) DLMgetras utility. Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL05437 The contents of the configuration Details...

  • Page 465

    Message Message Text Explanation devices, and then re-execute the The path (aa...aa) could not be dlmcfgmgr utility. deleted because the LU was in use. aa...aa: Path ID bb...bb: Error information (internal code) Action Stop any applications that are using the HDLM devices, and then re-execute the dlmcfgmgr utility for managing the HDLM configuration.

  • Page 466: Kapl06001 To Kapl07000, Kapl06001 To Kapl

    Message Message Text Explanation execute the command unless it is in You do not have a permission a boot script. required to execute dlmsetdevname. Action Do not execute the command unless it is in a boot script. KAPL05921 There is no symbolic link file for the Details HDLM nexus driver LinkGen has searched the...

  • Page 467

    Message Message Text Explanation information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM and report the error and detail code. For details on the DLMgetras utility, see DLMgetras Utility for Collecting HDLM Error Information on page 7-3.

  • Page 468

    Message Message Text Explanation information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM and report the error and detail code. For details on the DLMgetras utility, see DLMgetras Utility for Collecting HDLM Error Information on page 7-3.

  • Page 469: Kapl07001 To Kapl08000, Kapl07001 To Kapl

    Message Message Text Explanation aa...aa: Message code (hexadecimal number) bb...bb: Buffer size (hexadecimal number) Action Check whether any other error occurred. The information that could not be written is discarded. Review the actual memory size if another error does not occur. If the actual memory size is insufficient, increase the actual memory size.

  • Page 470: Kapl08001 To Kapl09000, Kapl08001 To Kapl

    KAPL08001 to KAPL09000 Message Message Text Explanation KAPL08019 The path (aa...aa) detected an error Details (bb...bb). (cc...cc) An error occurred in the path. The error is most likely due to a disconnected cable. aa...aa: Path identifier (hexadecimal number) bb...bb: Error code (hexadecimal number) When a path error is ¢...

  • Page 471

    Message Message Text Explanation ee...ee: Target ID (same as PathName of view -path) (hexadecimal number) ff...ff: HLU number (same as PathName of view -path) (hexadecimal number) gg...gg: Dev number (same as DNum of view -path) (decimal (base-10) number) hh...hh: Host device name (same as HDevName of view - path) Action...

  • Page 472: Kapl09001 To Kapl10000, Kapl09001 To Kapl

    Message Message Text Explanation aa...aa: Path ID (same as PathID of view -path) (decimal (base-10) number) Action Errors are detected in all the paths connected to the LUs. What To Do for a Path Error on page 5-3 to make the path shown in the error message or the paths connected to the target LU.

  • Page 473

    Message Message Text Explanation Delete unnecessary files to increase the amount of free capacity, and then retry. KAPL09008 The license code is invalid. Action Double-check the license code, and then re-enter it. KAPL09009 The license code is invalid. The Details HDLM installation program will The HDLM installation program is now terminate.

  • Page 474

    Message Message Text Explanation KAPL09021 An attempt to register the bundle Details PP name of Hitachi Network An attempt to register the PP Objectplaza Trace Library 2 failed. name of HNTRLib 2 has failed. Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.

  • Page 475

    Message Message Text Explanation /opt/HDVM/agent/docroot/ webstart/hdlm /opt/HDVM/agent/docroot/ hdlmhelp /opt/HDVM/agent/classes/jp /opt/HDVM/agent/classes/com/ Hitachi/soft/HiCommand/DVM/ agent/module/hdlm KAPL09026 Hitachi Network Objectplaza Trace Details Library 2 wasn't uninstalled Hitachi Network Objectplaza Trace because it was being used for Library 2 was not uninstalled other products. because it was being used for other products.

  • Page 476

    Message Message Text Explanation aa...aa: Year (4 digits)/month (01-12)/day (01-31) Action Install a permanent license by the expiration day. KAPL09078 The emergency license was Details installed. The license expires on An emergency license was aa...aa. installed. aa...aa: Year (4 digits)/month (01-12)/day (01-31) Action Install a permanent license by the...

  • Page 477

    Message Message Text Explanation Obtain a valid license key, and then re-install HDLM. KAPL09088 The entered license key is invalid. Action The HDLM installation will now Obtain a valid license key, and terminate. Obtain a valid license then re-install HDLM. key, and then re-install HDLM.

  • Page 478

    Message Message Text Explanation Contact your HDLM vendor or the A part of the HDLM configuration maintenance company if there is a file is missing. maintenance contract of HDLM. Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract of HDLM.

  • Page 479

    Message Message Text Explanation Action If a license key file exists, delete /var/tmp/hdlm_license Store a correct license key file in the designated directory, and then re-execute. KAPL09133 One or more required patches Details (aa...aa) could not be found. aa...aa: the number of the patch Please apply all required patches not yet applied then install again.

  • Page 480

    Message Message Text Explanation KAPL09140 HDLM cannot be installed in this Details zone. Install it in the global zone. HDLM cannot be installed in a non- global zone. Install HDLM in a global zone. Action Install HDLM in a global zone. KAPL09142 HDLM aa...aa cannot be Details...

  • Page 481

    Message Message Text Explanation Action Execute the DLMgetras utility for collecting HDLM error information to obtain the error information. Contact your HDLM vendor or your maintenance company if you have a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page...

  • Page 482

    Message Message Text Explanation /opt/DynamicLinkManager/ ¢ config/dlmfdrv.unconf.bak bb...bb: Internal code (decimal number) Action Execute the DLMgetras utility for collecting HDLM error information to obtain the error information. Contact your HDLM vendor or your maintenance company if you have a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility...

  • Page 483

    Message Message Text Explanation Restart the host with the reconfiguration option specified. KAPL09179 Data for maintenance: aa...aa Details bb...bb aa...aa: Message output location information (decimal (base-10) number) bb...bb: Detailed information (character string) Action None. KAPL09183 HDLM version aa...aa is installed. Details This version will now be aa...aa: Installed version number...

  • Page 484

    Message Message Text Explanation KAPL09191 The installation information Details settings file does not exist. The installation information settings file specified for the second parameter in the installhdlm utility for installing HDLM does not exist. Action Make sure that the path name of the installation-information settings file is appropriate, and then try again.

  • Page 485

    Message Message Text Explanation Check the absolute path of the license key file, and then re-install HDLM. KAPL09206 The entered value is invalid. Re- Details enter [1/2]: The entered value is invalid. Re- enter the value. Action Enter 1 or 2. KAPL09207 The entered absolute path is Action...

  • Page 486

    Message Message Text Explanation and then perform the action indicated in that error message. KAPL09213 An error occurred during aa...aa Details processing. Although the aa...aa processing has ended, an error occurred during the processing. aa...aa: installhdlm, pkgadd, dlnkmgr, or dlmsetconf Action Check the error message that was output just before this message,...

  • Page 487

    Message Message Text Explanation permissions for the file system of the /var/tmp directory. If the error code is 7: There is not enough free capacity for the file system of the /var/tmp directory. Make sure there is enough free capacity in the /var/tmp directory.

  • Page 488

    Message Message Text Explanation aa...aa: Error number (decimal number) that specifies the executed processing bb...bb: Return value (decimal number) of the executed processing Action If the error code is 9001,-1: You do not have read permissions for the installation information settings file. Make sure that you have access permissions.

  • Page 489

    Message Message Text Explanation retry the operation. If you do not have a required permission, grant required permissions to yourself, and then retry the operation. KAPL09219 An internal error occurred in the Details installhdlm_analysis. Error Code = An internal error occurred in the aa...aa,bb...bb installhdlm_analysis.

  • Page 490

    Message Message Text Explanation Specify the revised installation information settings file, and then retry the operation. KAPL09227 The definition of the installation Details information settings file includes A key that does not exist is an invalid key. Error Code = included.

  • Page 491

    Message Message Text Explanation Specify the revised installation information settings file and then retry the operation. KAPL09230 The definition of the installation Details information settings file includes a A duplicated section name is duplicated section name. Error included. Code = aa...aa,bb...bb, line = aa...aa: Error number (decimal cc...cc number) that specifies the...

  • Page 492

    Message Message Text Explanation Specify the revised installation information settings file and then retry the operation. KAPL09233 The definition of the installation Details information settings file is too The length of a single line of the long. Error Code = aa...aa,bb...bb, definition exceeds 1023 line = cc...cc characters.

  • Page 493

    Message Message Text Explanation necessary, and then retry the operation. If the error code is 3: You do not have write permission for the file system of the log file. Modify your access permissions as necessary, and then try again. If the error code is 4: You do not have write permission for the...

  • Page 494

    Message Message Text Explanation A file other than a normal one (for example, a directory file or a special file.) is specified. aa...aa: Specified file name (character string) Action Specify the correct file, and then retry the operation. KAPL09239 The system will now restart. Details The host will restart because restart was specified in the...

  • Page 495

    Message Message Text Explanation Action Install the prerequisite JDK, and then execute the HDLM component installation utility (dlminstcomp). KAPL09243 dlminstcomp completed Details successfully. The HDLM component installation utility (dlminstcomp) finished successfully. Action None. KAPL09246 An attempt to install an HDLM Details component failed because an An attempt to install an HDLM...

  • Page 496

    Message Message Text Explanation Action Check whether HDLM has been correctly installed. KAPL09504 The language environments of Details HDLM and the Service Pack are The Japanese SP was applied to different. the English edition of HDLM, or vice versa. Action Acquire the SP that has the same language as the installed HDLM, and then try again.

  • Page 497

    Message Message Text Explanation aa...aa: The version of the SP to be installed bb...bb: The version of the installed HDLM or the version of the installed SP Action You do not have to install the SP. Keep using HDLM as is. KAPL09601 Cannot install in this system.

  • Page 498: Kapl10001 To Kapl11000, Kapl10001 To Kapl

    Message Message Text Explanation KAPL09604 A system error occurred. The Details installation will now stop. aa...aa: Command that failed (command = aa...aa, error code = bb...bb: Return value bb...bb) Action Acquire the log /var/tmp/ hdlm_installux_sh.log, and then contact your HDLM vendor or, if you have a maintenance contract, contact the maintenance company.

  • Page 499

    Message Message Text Explanation Check the parameters of the DLMgetras utility for collecting HDLM error information, and then retry. For details on the DLMgetras utility, see DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10004 The parameter contains an Details incorrect value.

  • Page 500

    Message Message Text Explanation KAPL10007 A directory has been specified in Details the third parameter. Value = The -f parameter is specified to a aa...aa directory. aa...aa: Third parameter Action Check the parameters of the DLMgetras utility for collecting HDLM error information, and then retry.

  • Page 501

    Message Message Text Explanation file for defining the information to be collected. aa...aa: Line number of the file for defining information to be collected (decimal number) Action Delete the coding of the root directory from the specified file. The displayed directory will be ignored and the DLMgetras utility for collecting HDLM error information will continue.

  • Page 502

    Message Message Text Explanation be collected (decimal (base-10) number) bb...bb: Indicated contents of a line Action After the DLMgetras utility for collecting HDLM error information terminates, check the contents of the file for defining which information to be collected. This file is shown in the message.

  • Page 503

    Message Message Text Explanation defining information to be collected. aa...aa: Line number of the file for defining information to be collected (decimal number) bb...bb: Indicated contents in a line Action After the DLMgetras utility for collecting HDLM error information terminates, check the contents of the file for defining the information to be collected.

  • Page 504

    Message Message Text Explanation HDLM Error Information on page 7-3. KAPL10017 You lack privileges for executing Details the utility for collecting HDLM error The DLMgetras utility for information. collecting HDLM error information must be executed by a user with root permissions. Action Re-execute as a user with root permissions.

  • Page 505

    Message Message Text Explanation because the user replied to the confirmation with an n response. Action None. For details on the DLMgetras utility, see DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10031 The entered value is invalid. Details Continue operation ? [y/n]: A value other than y or n has...

  • Page 506

    Message Message Text Explanation KAPL10035 An attempt to archive the error Details information failed. Details = aa...aa Execution of the tar command failed. aa...aa: tar command output message Action See the details in the message, and then remove the cause of the error.

  • Page 507

    Message Message Text Explanation None. KAPL10201 usage: dlmsetconf [-d output- Details directory-name] [-r|-u] Displays how to use dlmsetconf. Action Specify appropriate parameters, and then retry. KAPL10202 A backup file (aa...aa) could not be Details created. Errno = bb...bb (cc...cc) An attempt to create a backup file has failed.

  • Page 508

    Message Message Text Explanation Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM if this error occurs a number of times. KAPL10224 An error occurred when creating Details the configuration definition An attempt to create a information.

  • Page 509

    Message Message Text Explanation An attempt to open a directory has failed. aa...aa: Directory name bb...bb: Errno (decimal number) cc...cc: Errno string Action Remove the error by referencing Errno, and then retry. KAPL10231 The aa...aa file could not be Details deleted.

  • Page 510

    Message Message Text Explanation Allocate the necessary amount of memory by freeing up unnecessary resources. KAPL10237 The property information (aa...aa) Details on a device (bb...bb) could not be An attempt to acquire property collected. information for the sd or ssd device has failed.

  • Page 511

    Message Message Text Explanation device files of the HDLM-managed sd or ssd device are deleted at reboot. Action After you execute the dlmsetconf utility, the logical device files of the HDLM-managed sd or ssd device are deleted at reboot. To configure HDLM, enter y if you want to execute the dlmsetconf utility, enter n to stop.

  • Page 512

    Message Message Text Explanation An attempt to acquire link destination information about the symbolic link file has failed. aa...aa: File name bb...bb: Errno (decimal number) cc...cc: Errno string Action See errno to eliminate the error, and then retry. KAPL10253 An attempt to create symbolic link Details aa...aa to a file failed.

  • Page 513

    Message Message Text Explanation KAPL10257 The output file (aa...aa) could not Details be opened. Errno = bb...bb An attempt to open the output file (cc...cc) has failed. aa...aa: Output file name bb...bb: Errno (decimal number) cc...cc: Errno string Action Take countermeasures according to the Errno.

  • Page 514

    Message Message Text Explanation KAPL10260 The LDEV property was assumed Details because the storage model ID The LDEV property was assumed could not be acquired. aa...aa because the storage model ID could not be acquired. aa...aa: error information Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.

  • Page 515

    Message Message Text Explanation Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL10320 The file format is invalid. File name Details = aa...aa The file format is invalid. aa...aa: File name Action Re-execute the HDLM- configuration definition utility.

  • Page 516

    Message Message Text Explanation Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL10368 usage: dlmcfgmgr {-a|-h} Details Displays how to use dlmcfgmgr. Action Specify appropriate parameters, and then retry. For details on the dlmcfgmgr utility, see dlmcfgmgr Utility for Managing the HDLM Configuration on page...

  • Page 517

    Message Message Text Explanation Action Remove the error by referencing ErrorCode(errno), and then retry. KAPL10373 This operation will change the Details configuration of HDLM devices, and This message confirms the delete device files of sd/ssd devices execution of the dlmcfgmgr utility managed by HDLM.Do you want to for managing the HDLM continue? [y/n]:...

  • Page 518

    Message Message Text Explanation Take action in accordance with ErrorCode (errno), and then retry. KAPL10377 The HDLM device configurations Details have not been changed. There was no change in the HDLM device configuration. Action None. KAPL10378 The LU (LDEV = aa...aa) was Details successfully added.

  • Page 519

    Message Message Text Explanation KAPL10383 The text aa...aa could not be found Details in the existing configuration The property name of the search definition file (bb...bb). Execute the string does not exist in the / DLMgetras utility to collect error kernel/drv/dlmfdrv.conf file.

  • Page 520

    Message Message Text Explanation bb...bb: Property name cc...cc: LDEV name Action If you want to use the LU, reboot the host. KAPL10388 The path (aa...aa) could not be Details added. (Function: bb...bb, Error An attempt to add a path has Code = cc...cc (dd...dd)).

  • Page 521

    Message Message Text Explanation cc...cc: Errno string Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL10395 Multiple paths of the LU (LDEV = Details aa...aa) to be added have the same The same property value exists value as the property (bb...bb).

  • Page 522

    Message Message Text Explanation Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL10701 usage: dlmvxexclude [ -d dir | -h ] Details Displays how to use dlmvxexclude. Action Specify appropriate parameters, and then retry.

  • Page 523

    Message Message Text Explanation Creation of the VxVM Configuration File on page 7-26. KAPL10705 Internal processing error. Details An internal error occurred while the dlmvxexclude utility for assisting creation of the VxVM configuration file was executing. Action Check the system configuration and then retry.

  • Page 524

    Message Message Text Explanation aa...aa: File name Action Check the system configuration. KAPL10710 The existing configuration definition Details file aa...aa could not be opened. An attempt to open a Errno = bb...bb (cc...cc) configuration definition file has failed. aa...aa: Name of the configuration definition file bb...bb: Errno (decimal number) cc...cc: Errno string...

  • Page 525

    Message Message Text Explanation KAPL10714 Could not allocate memory. Errno Details = aa...aa (bb...bb). An attempt to reserve memory has failed. aa...aa: Errno (decimal number) bb...bb: Errno string Action Check if more than necessary memory has been allocated. If so, release the unnecessary resources.

  • Page 526

    Message Message Text Explanation Take countermeasure according to the Errno and then execute the command again. KAPL10719 The file aa...aa already exists. Do Details you want to overwrite it? [y/n]: This message confirms the execution of the dlmvxexclude utility for assisting creation of the VxVM configuration file.

  • Page 527

    Message Message Text Explanation None. For details on the dlmvxexclude utility, see dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-26. KAPL10723 The execution of dlmvxexclude is Details not required for the configuration. The devices from VxVM do not need to be disabled so the dlmvxexclude utility does not need not be executed.

  • Page 528

    Message Message Text Explanation KAPL10754 A user terminated the dlmsetboot Details utility. End time = aa...aa Processing of the dlmsetboot utility has been terminated because the user replied to the confirmation with an n response. aa...aa: year (4 digits)/month/ date hour:minute:second Action None.

  • Page 529

    Message Message Text Explanation KAPL10758 You lack privileges for executing Details the dlmsetboot utility. The dlmsetboot utility must be executed by a user with root permissions. Action Re-execute as a user with root permissions. For details on the dlmsetboot utility, see dlmsetboot Utility for Assisting Configuration of an HDLM Boot Disk Environment on page...

  • Page 530

    Message Message Text Explanation Specify a disk other than the pre- migration boot disk device for the post-migration boot disk device. KAPL10763 The partitions on the post- Details migration boot disk device are The partitions on the post- different from those on the pre- migration boot disk device are migration boot disk device.

  • Page 531

    Message Message Text Explanation To manually migrate Sun Cluster global device file systems or SVM state databases after the utility ends, enter y. To end the utility and have the user manually perform the migration by following the procedure in Configuring a Boot Disk Environment on page 3-76, enter...

  • Page 532

    Message Message Text Explanation aa...aa: boot disk slice after migration bb...bb: partition size (bytes) cc...cc: boot disk slice before migration dd...dd: partition size (bytes) Action Recheck the partition size on the post-migration boot disk device. KAPL10770 This version of the OS is not Details supported.

  • Page 533: Kapl11001 To Kapl12000, Kapl11001 To Kapl

    Message Message Text Explanation (LDEV = bb...bb) that will be Multiple paths have the same deleted. The LU configuration will property value for a LU registered not be changed. in the filter driver information. aa...aa: Property name = property value bb...bb: LDEV name Action Contact your HDLM vendor or the...

  • Page 534

    Message Message Text Explanation bb...bb: The Path ID of the target path (decimal number) Action None. KAPL11903 aa...aa has completed normally. Details The operation has completed normally on the host. aa...aa: Any of the following operations (character string) Get Path Information ¢...

  • Page 535

    Message Message Text Explanation Get Path Status Log ¢ Information bb...bb: Error status returned from API (character string) Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if you have a maintenance contract for HDLM.

  • Page 536: Kapl13001 To Kapl14000, Kapl13001 To Kapl

    Message Message Text Explanation aa...aa: XML information (character string) Action None. KAPL13001 to KAPL14000 Message Message Text Explanation KAPL13601 The audit log configuration file does Details not exist. Restart the HDLM The audit log configuration file Manager, and execute the "dlnkmgr does not exist.

  • Page 537: Kapl15001 To Kapl16000, Kapl15001 To Kapl

    Message Message Text Explanation KAPL13604 An error occurred during processing Details to read the audit log configuration An internal error occurred during file. processing to read the audit log configuration file. Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.

  • Page 538

    Message Message Text Explanation KAPL15023 An attempt to create the Details configuration file for the HDLM driver aa...aa: Name of the utility that has failed. Command Line = aa...aa the user executed KAPL15024 The HDLM-management target was Details successfully set. Command Line = aa...aa: Name of the utility that aa...aa the user executed...

  • Page 539

    Message Message Text Explanation cc...cc: Command that the user executed KAPL15105 Setting up the operating Details environment succeeded. Command aa...aa: Command that the user Line = aa...aa executed KAPL15106 Setting up the operating Details environment failed. Command Line = aa...aa: Command that the user aa...aa executed KAPL15107...

  • Page 540: Return Codes For Hitachi Command Suite Common Agent Component

    Message Message Text Explanation KAPL15404 The HDLM Manager was executed by the user who does not have the authority. Return Codes for Hitachi Command Suite Common Agent Component When an operation requested of HDLM from Global Link Manager terminates abnormally, or terminates normally with a warning, HDLM outputs one of the return codes described below.

  • Page 541

    Message ID Explanation Action Remove the error in the path, and then retry. 1016 Details The target path(s) are already Online. Action Update the host information, and then check the path status. 1017 Details The target path(s) are already Offline(C). Action Update the host information, and then check the path status.

  • Page 542

    Message ID Explanation Refresh the host information, check the path status, and then perform the operation again. 1027 Details The error monitoring interval and the number of times that the error needs to occur are conflicting with the automatic failback-checking interval. Action Set the intermittent error-monitoring interval to a value that is equal to or greater than (automatic-failback-checking-...

  • Page 543

    Message ID Explanation An attempt to recover the path failed because an error occurred on the host. Action Re-execute. If the same error occurs even after removing the reservation, execute the DLMgetras utility for collecting HDLM error information to collect the error information, and then contact your HDLM Refresh the host information, and then perform the operation again.

  • Page 544

    Message ID Explanation Re-execute the command. If the same error occurs again, execute the DLMgetras utility for collecting HDLM error information to collect the error information, and then contact your HDLM vendor or maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page...

  • Page 545

    Message ID Explanation execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or maintenance company if you have a maintenance contract for HDLM. For details on the DLMgetras utility, see DLMgetras Utility for Collecting HDLM Error Information on page 7-3.

  • Page 546

    8-122 Messages Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 547

    Sun Cluster 3.2 Commands This section describes the Sun Cluster 3.2 commands. □ Sun Cluster 3.2 Commands Sun Cluster 3.2 Commands Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 548

    Sun Cluster 3.2 Commands Table A-1 Correspondence Between Sun Cluster 3.1 Commands and Sun Cluster 3.2 Commands on page A-2 shows the correspondence between Sun Cluster 3.1 commands and Sun Cluster 3.2 commands. For details about the command options, see the documentation for Sun Cluster 3.2. Note that there are no differences between Sun Cluster 3.2 commands and Oracle Solaris Cluster 3.3 commands.

  • Page 549

    Commands for Sun Cluster 3.1 Commands for Sun Cluster 3.2 scswitch -F -D device-group-name cldevicegroup offline device- group-name scswitch -S -h node-name clnode evacuate node-name Execute this command in step 13 in Set Up Sun Cluster on page 3-47 Preparations for a New Installation of HDLM on page 3-41 and in step 14 Remove Sun Cluster Settings on page 3-166 Remove Sun Cluster...

  • Page 550

    Sun Cluster 3.2 Commands Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 551: Functional Differences Between Version 6.1 Or Later And Versions Earlier Than 6.1

    Functional Differences Between Versions of HDLM This section explains the functional differences between newer and older versions of HDLM. □ Functional Differences Between Version 6.1 or Later and Versions Earlier Than 6.1 □ Functional Differences Between Version 6.0 or Later and Versions Earlier Than 6.0 □...

  • Page 552

    Functional Differences Between Version 6.1 or Later and Versions Earlier Than 6.1 In a new installation of HDLM, the default load balancing algorithm has been changed from the Round Robin algorithm to the Extended Least I/Os algorithm. Functional Differences Between Version 6.0 or Later and Versions Earlier Than 6.0 •...

  • Page 553: Acronyms And Abbreviations

    Acronyms and abbreviations The following acronyms and abbreviations might be used in this guide. Arbitrated Loop Application Programming Interface Automatic Storage Management Channel Adapter CLPR Cache Logical Partition Central Processing Unit Control Unit Acronyms-1 Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 554

    DBMS Database Management System Device Dynamic Multipathing Domain Name Server Dynamic Reconfiguration Fibre Channel FC-SP Fibre Channel Security Protocol Failover FQDN Fully Qualified Domain Name Group Membership and Atomic Broadcast Greenwich Mean Time Graphical User Interface Acronyms-2 Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 555

    Host Bus Adapter HDev Host Device Host Logical Unit HTTP Hypertext Transfer Protocol Input/Output Internet Protocol iSCSI Internet Small Computer System Interface Local Area Network LDAP Lightweight Directory Access Protocol LDEV Logical Device LDoms Logical Domains Low Latency Transport Acronyms-3 Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 556

    Logical Unit Logical Unit Number Maintenance Pack Network Attached Storage Network File System Network Time Protocol OpenBoot PROM Operating System Port RADIUS Remote Authentication Dial in User Service Acronyms-4 Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 557

    Storage Area Network SCSI Small Computer System Interface SLPR Storage Logical Partition SMTP Simple Mail Transfer Protocol SNMP Simple Network Management Protocol Service Pack Secure Sockets Layer Service Processor UNIX File System VTOC Volume Table Of Contents World Wide Name Acronyms-5 Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 558

    Zettabyte File System Acronyms-6 Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 559: Automatic Failback

    Glossary This glossary explains the terms used in this manual. automatic failback A function for checking the status of paths at regular intervals, and automatically placing the status of a path recovered from an error into the Online status. If a path with an Offline(E) or Online(E) status recovers from an error, an automatic failback will place the path Online.

  • Page 560

    CLPR(Cache Logical Partition) A function supported by the Hitachi USP, Universal Storage Platform V/VM, Virtual Storage Platform, and Hitachi AMS2000/AMS/WMS/SMS series for logically splitting up a cache. This function can split up a cache into parity groups in the storage system, so that other parity groups do not affect the cache performance.

  • Page 561

    failback A function for placing the status of a path recovered from an error into the Online status, and then switching the access path to the path that was just recovered. failover A function for switching to another normal path if there is an error in the current access path, thereby enabling the system to continue to operate.

  • Page 562: Load Balancing

    host device name A name assigned to a host device. A name of the logical device file for an HDLM device is assigned to a host device name. A host device name is assigned using the cUtXdY format. The variables used in cUtXdY indicate the following: U: Controller number reserved by HDLM X: Target ID or World Wide Name of the sd or ssd device for the HDLM device Y: LUN of the sd or ssd device for the HDLM device...

  • Page 563: Logical Device File

    logical device file Device files under /dev/dsk and /dev/rdsk. Note that the term logical device used here is not an LDEV. LU (Logical Unit) A logical unit that is a logical volume defined in the storage system, and with which the host can perform input or output operations.

  • Page 564: Physical Path

    path An access path from a host to a storage system. Access to an area within an LU in a storage system is made via a cable connecting the HBA on the host and the CHA on the storage system. This cable is a path. Each path has an AutoPATH_ID. (See also: AutoPATH_ID) path health checking A function for checking the status of paths at regular intervals.

  • Page 565

    systems, and prevents the transfer of high-volumes of data from deteriorating LAN performance. SCSI device A SCSI disk device SLPR (Storage Logical Partition) The SLPR function enables logical splitting of a storage system supported by the Hitachi USP, Universal Storage Platform V/VM and Virtual Storage Platform. This function splits up the resources, such as ports, CLPR, and volumes, in the storage system, so that the user can manage each resource independently.

  • Page 566

    V W X Glossary-8 Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 567

    Index Symbols buffer size 3-129 changing the buffer size per monitoring interval /etc/system file duration 3-132 setup 3-45 bus number 6-48 /kernel/drv/sd.conf file setup 3-44 CHA (Channel Adapter) Glossary-1 active host 2-40 changing path configuration adding in a boot disk environment 4-19 a new logical unit 4-22...

  • Page 568

    monitor 6-63 error levels 2-30 components Critical 2-30 HDLM system components Error 2-30 core logic component Information 2-30 Critical 2-30 Warning 2-30 cUtXdYsZ 2-10 error log 2-29 error log collection level 6-18 values 6-19 error path identifying default and recommended values 6-17 error-log-file-size 6-19...

  • Page 569

    path error help Hitachi Network Objectplaza Trace Library HBA (Host Bus Adapter) Glossary-3 applying the settings 3-134 hbsa.tar.Z Displaying setup menu 3-130 HDev Notes 3-130 HDevName Hitachi Network Objectplaza Trace Library HDLM (HNTRLib) features uninstalling 3-181 functions Hitachi Network Objectplaza Trace Library installing 3-41, 3-49, 3-56 (HNTRLib2)

  • Page 570

    installing HDLM LDoms environment 3-71 manual path switching 2-18 installux.sh 7-36 message integrated HDLM management 2-39 examining 5-5, 5-8 integrated trace file 2-29 message level changing the number 3-131 Messages changing the size 3-131 adjusting the number of messages to be output integrated trace information files 3-129 3-133...

  • Page 571

    HDLM owner controller 2-12, Glossary-5 owner path 2-12, Glossary-5 SAN (Storage Area Network) Glossary-6 SCSI device Glossary-7 setting up 3-156 SDS settings path 2-3, Glossary-6 Removing 3-169 checking the configuration 3-119 serial number 6-45 path error 6-16 handling setup path health checking 2-27, 6-20, Glossary-6 /etc/system file 3-45...

  • Page 572

    trace level values 6-20 VxVM (before installing HDLM) types of collected logs 2-28 setup 3-46 types of error information 2-29 VxVM volume creating 3-139 unattended installation installation-information settings file 7-29 Warning 2-30 log file 7-35 what is HDLM? preparation 3-49 what to do for uninstallation program error...

  • Page 573

    Hitachi Dynamic Link Manager User Guide (for Solaris(R))

  • Page 574

    Hitachi Data Systems Corporate Headquarters 750 Central Expressway Santa Clara, California 95050-2627 U.S.A. Phone: 1 408 970 1000 www.hds.com info@hds.com Asia Pacific and Americas 750 Central Expressway Santa Clara, California 95050-2627 U.S.A. Phone: 1 408 970 1000 info@hds.com Europe Headquarters Sefton Park Stoke Poges Buckinghamshire SL2 4HD...

Comments to this Manuals

Symbols: 0
Latest comments: