Download  Print this page

NEC Storage Performance Monitor/Optimizer User Manual

Hide thumbs

Advertisement

NEC Storage
Performance Monitor/Optimizer
User's Manual

Advertisement

Table of Contents
loading

  Summary of Contents for NEC Storage Performance Monitor/Optimizer

  • Page 1 NEC Storage Performance Monitor/Optimizer User’s Manual...
  • Page 2 © NEC Corporation 2001-2003 No part of the contents of this book may be reproduced or transmitted in any form without permission of NEC Corporation. The contents of this book may be modified without notice in the future.
  • Page 3 The performance optimization function is for optimizing the performance of the disk array by moving a logical disk in the disk array, displaying the Busy Ratio, and estimating the Busy Ratio after it has been moved. Refer to the “NEC Storage Manager Manual Guide” (IS901) for the overview of NEC Storage Manager and the related manuals.
  • Page 4 6. In this document, matters to which careful attention needs to be paid will be described as follows: Be sure to observe the contents. If the indications are ignored and the system is improperly operated, settings which have been already made might be affected.
  • Page 5: Table Of Contents

    Command for Performance Tuning .......................... 17 Appendix B CSV File Structure............................20 Appendix C Report Format of Performance Report Editor .................... 36 Appendix D Estimation of NEC Storage PerformanceMonitor File Size ............... 46 Appendix E Estimating Size of NEC Storage PerformanceOptimizer Files..............48...
  • Page 6 Appendix F Example of Performance Improvement......................49 Index ................................... 51...
  • Page 8: Part I Performance Monitoring Function

    The Performance Monitoring Function provides a performance monitoring function for centrally controlling the storage performance under the control of NEC Storage Manager (hereafter, referred to as iSM). The use of this performance monitoring function is capable of real-time monitoring of loads, detection of overload, and management of statistic information (performance information of the past).
  • Page 9: System Configuration

    NEC Storage PerformanceMonitor requires the following program product as a precondition. • NEC Storage BaseProduct Refer to Chapter 3 “System Configuration of NEC Storage Manager” of the “NEC Storage Manager Manual Guide”. This section describes operations of the Performance Monitoring Function. The Performance Monitoring Function is...
  • Page 10 Chapter 1 Overview of Performance Monitoring Function Reviewing cause of overload Displaying statistic information on a real-time basis Display performance information on the real-time basis and check 1.5.1 Real-Time Display of Load State location of overload. Then, analyze the accumulated statistic information to determine the cause of the overload.
  • Page 11 For details of components composing disk array, refer to 3.3 “Hardware” in the “NEC Storage Manager Manual Guide”. The iSM client screen is capable of displaying the index values for each calculation unit shown in Table 1-1. (Index values applicable to the items indicated by “...
  • Page 12 Chapter 1 Overview of Performance Monitoring Function Displays the latest load state in bar graphs Figure 1-3 Displaying the Bar Graph...
  • Page 13 Chapter 1 Overview of Performance Monitoring Function Displays the latest load state in time-series graphs Figure 1-4 Displaying the Time-series Graph...
  • Page 14 Chapter 1 Overview of Performance Monitoring Function (1) Monitored Items Monitored items are types of index values which indicate status of load or performance. iSM displays values of the following types of monitored items on the client screen. External I/O Density This is the number of input/output requests received by the disk array from the external per unit time.
  • Page 15 Chapter 1 Overview of Performance Monitoring Function data transfer during the unit measurement period and dividing it by the time interval of the measurement period. Internal Average Transfer Length This is the average input/output data length of the input/output process performed for the internal disk devices by the control device of the disk array.
  • Page 16 Monitor screen. When occurrence of an overload is reported by a message, an administrator must identify location of the overload on the Performance Monitor screen. Business Server Performance Monitoring Screen ARRAY1 ARRAY2 NEC Storage Overload PerformanceMonitor Overload Occurrence Display ARRAY1 ARRAY2 Operations Administrator NEC Storage Series I-10...
  • Page 17 (Note 1) L2 or L3. (Note 1) For user levels, refer to 5.2 “Client Start/Stop” of the “NEC Storage Manager User’s Manual” of your OS. (1) Notifying threshold excess by e-mail You can also e-mail the message notifying you of occurrence of an exceeded threshold to a specified address by using the link function of the iSM server.
  • Page 18 Chapter 1 Overview of Performance Monitoring Function しきい値監視インターバルが短いため、 一時的な Because of short threshold monitoring interval, exceeded 過負荷でしきい値超えが通知される threshold is notified upon a temporary overload. value of 入出力回数 monitored item Threshold value しきい値 Specifying “1 minute” for a しきい値監視インターバルを threshold monitoring interval 1 分と設定した場合...
  • Page 19 Chapter 1 Overview of Performance Monitoring Function Detect threshold し きい値超え検出 し きい値超え検出 し きい値超え検出 し きい値超え検出 exceeding Threshold monitoring し きい値監視インタ ーバル interval し きい値超え状態 し きい値超え状態 し きい値超え状態 し きい値超え状態 Threshold excess state し きい値監視インタ ーバル( 3 回) し...
  • Page 20 1. Accumulating statistic information 2. Concentrating, extracting, and editing statistic information Statistic information NEC Storage 2. Downloading history file accumulated statistic PerformanceMonitor Storage information 3.
  • Page 21 PerformanceMonitor ARRAY1 ARRAY2 Statistic Information History File NEC Storage Series Figure 1-9 Accumulating Statistic Information Statistic information history files are created based on the following naming conversion. Naming Convention for Statistic Information History Files: yyyymmdd Disk Array Name [#nnn].psl yyyymmdd The year, month, and day when accumulation to the statistic information history file started.
  • Page 22 For information on how to estimate the size of a statistic information history file, refer to Appendix D “Estimation of NEC Storage PerformanceMonitor File Size”. The period of reading statistic information is called the logging Interval. The user can specify the logging Interval from 1 minute to 60 minutes at environment definition of the iSM server.
  • Page 23 Chapter 1 Overview of Performance Monitoring Function Table 1-3 Types of Statistic Information Accumulated in Statistic Information History Files Unit of Summarizing By Logical Disk By Host Port By Disk Port (Note1) By Physical Disk By RANK Write External I/O Count External I/O Count Internal I/O Count Internal I/O Count...
  • Page 24 Chapter 1 Overview of Performance Monitoring Function RANK is not accumulated in a statistic information history file. I-18...
  • Page 25 Chapter 1 Overview of Performance Monitoring Function (1) Types of statistic information iSM accumulates the following types of statistic information in statistic information history files. External I/O count This is the number of input/output requests received by the disk array from the external in each measurement period. The unit is the number of times.
  • Page 26 Chapter 1 Overview of Performance Monitoring Function millisecond. (2) Unit to sum up Units to sum up are the units used for summarizing statistic information. iSM summarizes statistic information by the following types of components and accumulates them separately. By logical disk Load state of each logical disk of the monitored disk array is accumulated.
  • Page 27 For information on how to estimate the size of a statistic information summarized file, refer to Appendix D “Estimation of NEC Storage PerformanceMonitor File Size” of this manual.
  • Page 28 Chapter 1 Overview of Performance Monitoring Function Maximum/minimum This is the maximum and minimum values of the external I/O Density in each External I/O Density measurement period contained in the period of summarizing. The unit is count/1000 seconds. Maximum/minimum This is the maximum and minimum values of the external transfer rates in each External Transfer Rate measurement period contained in the period of summarizing.
  • Page 29 Chapter 1 Overview of Performance Monitoring Function History file Primary summarized file (primary summarizing) The input is a statistic information history file, and the output is a primary summarized file. In other words, statistic information in each logging Interval within each period of summarizing in the input file is summarized in units of hours and the result is recorded in the output file.
  • Page 30 Chapter 1 Overview of Performance Monitoring Function specified for CSV Conversion Tool. Period of extraction Component of disk array Type of statistic information In statistic information history/summarized files, the component information of the disk array is accumulated in addition to statistic information.
  • Page 31 Chapter 1 Overview of Performance Monitoring Function (1) Extracting statistic information in the specified period To extract statistic information in the specified period, specify the period in CSV Conversion Tool. To specify a period, you can specify the extraction starting time and end time, or either one of them. Specifying the extraction starting time and end time Statistic information from the extraction starting time to the extraction end time is extracted.
  • Page 32 Chapter 1 Overview of Performance Monitoring Function Disk port Logical Disk Physical Disk RANK (ii) Information regarding changes in logical disk names This is information regarding changes in logical disk names. (iii) Information regarding changes in disk array configuration This is information regarding changes in disk array configuration. (iv) Information regarding changes in RANK configuration This is information regarding Logical Disk moving between RANKs by the performance optimization function.
  • Page 33 Chapter 1 Overview of Performance Monitoring Function host director ports disk director ports logical disks physical disks RANKs Analysis of the Performance Summary Report provides understanding of general load status in the entire disk array and each component of the statistic information collection period. (ii) Performance Detail Report The Performance Detail Report summarizes the following statistic information per statistic information logging...
  • Page 34 Chapter 1 Overview of Performance Monitoring Function The Performance Report Viewer is a tool that uses detailed report on the whole disk arrays created via the Performance Report Editor as inputs to provide graphic representation of statistic information. The user references a graph to understand the load state of disk array and chronological change in the load.
  • Page 35: Server Installation

    Installation of iSM is accompanied by installing the NEC Storage PerformanceMonitor at the same time. For installing and uninstalling, refer to Part I “Installation and Setting of NEC Storage Manager” of the “NEC Storage Manager User’s Manual” for your OS.
  • Page 36: Client Installation

    The same system requirements as those for iSM are applicable to the NEC Storage PerformanceMonitor. Refer to Part I “Installation and Setting of NEC Storage Manager” of the “NEC Storage Manager User’s Manual” for your OS. For installing the NEC Storage PerformanceMonitor, install the iSM client with specifying a setup type that includes extended functions.
  • Page 37 Chapter 3 Operations of Performance Monitoring Function Chapter 3 Operations of Performance Monitoring Function Explanation will be made about the operation method for monitoring the load status in real time by using the Performance screen. (1) Performance Monitoring Starting Clicking the button for activating the performance monitor function opens the Performance Monitor screen, which starts performance monitoring.
  • Page 38 Chapter 3 Operations of Performance Monitoring Function The performance monitor screen, as shown in Figure 3-1, is composed of areas (i) thorough (v). The following describes the areas. Menu bar The menu bar is used for selecting necessary items from the menu for various operations. For details, refer to Help.
  • Page 39 Chapter 3 Operations of Performance Monitoring Function Indicates that iSM has not completed in obtaining statistic information required for displaying client. (Note 1) A performance monitor server using older version software may not support some of the displaying functions. In such a case, a hyphen “-”...
  • Page 40 Chapter 3 Operations of Performance Monitoring Function External Transfer Rate External Average Transfer Length External Average Response Time However, the string “External” is omitted in the labels. When you change the disk array to be monitored by using the combo box, the display switches to the list of load status by port of the selected disk array.
  • Page 41 Chapter 3 Operations of Performance Monitoring Function (4) Registering/Deleting logical disks to be monitored When you click [Setting] and then select [Monitoring Object Logical Disk] on the Numeric Table screen (Figure 3-2), the Registration of Logical Disks dialog box appears (Figure 3-3). This dialog allows you to change the logical disks for performance monitoring in the disk array specified in the combo box of the Numeric Table.
  • Page 42 Chapter 3 Operations of Performance Monitoring Function Figure 3-4 Registration of Physical Disks Dialog [Monitoring Candidate List] displays unregistered physical disks among all the physical disks created in the specified disk array. [Monitor Object] displays all the physical disks currently registered as monitoring objects. Selecting physical disks in the [Monitoring Candidate List] and clicking the [Add] button will register the specified physical disks to the [Monitor Object] list.
  • Page 43 Chapter 3 Operations of Performance Monitoring Function Figure 3-5 Registering to Graphs Menu On the Numeric Table screen, when you specify a line which shows the latest value (place the cursor over the left cell and click the cell. In Figure 3-5, the line of logical disk 008ch is specified), click [Setting], and then point to [Registration to Graph], all the monitored items for the line are registered as items to be displayed in Time-series Graphs.
  • Page 44 Chapter 3 Operations of Performance Monitoring Function Figure 3-6 Time-series Graph Screen The “Time-series Graph” screen consists of the menu area and tool bar for performing various operations, the load status display area for showing load status, the legend display area for showing the items displayed in Time-series Graphs, and the status bar for displaying the connection state.
  • Page 45 Chapter 3 Operations of Performance Monitoring Function The threshold line is displayed on the graph. The line is displayed for time zone from the time the threshold is set to the time it is unset. When past information is displayed, a threshold line is displayed in the time zone for which thresholds have been set.
  • Page 46 Chapter 3 Operations of Performance Monitoring Function (Note 1) A performance monitor server using older version software may not support some of the displaying functions. In such a case, a graph frame only is displayed and graph lines are not displayed. For displaying graph lines, software of the latest version is required.
  • Page 47 Chapter 3 Operations of Performance Monitoring Function Average Response Time Read Hit Ratio Write Hit Ratio Busy Ratio This section describes the procedures for displaying a line indicating a threshold (index value for overloads) on the graph. For threshold, refer to 3.2 “Monitoring Overloads”. On the Bar Graph screen, select a line or multiple lines for which you want to display threshold from the legend display area, and then select [View] [Show Threshold in graph].
  • Page 48 Chapter 3 Operations of Performance Monitoring Function (9) Changing Graph Scales A maximum value of a graph scale displayed on the Bar Graph screen is changeable freely on the Setting Graph Scale dialog. The Setting Graph Scale dialog is displayed by selecting [Setting] [Set Graph Scale] menu.
  • Page 49 Chapter 3 Operations of Performance Monitoring Function (11) Saving Settings To save registration and deletion settings for graphs performed in the [Setting] menu and change settings of logical/physical disks to be monitored, click [File], and then point to [Save of Setting] as shown in Figure 3-10. The saved settings are restored when the Performance Monitoring screen opens next time.
  • Page 50 Chapter 3 Operations of Performance Monitoring Function (12) Performance Monitoring Stop To close the Performance Monitoring screen, click [File] and then click [Close] Menu as shown in Figure 3-11. Figure 3-11 Close Menu I-44...
  • Page 51 Chapter 3 Operations of Performance Monitoring Function (1) Thresholds display When you click [View] and then click [Threshold] on the Numeric Table screen as shown in Figure 3-12, the threshold column appears next to the current value column, as shown in Figure 3-13, to display thresholds of the monitored items specified for each monitoring target.
  • Page 52 Chapter 3 Operations of Performance Monitoring Function (2) Thresholds setting On the Numeric Table screen, select the line (Note 1) or cell (Note 2) of the threshold monitoring element for which you want to set a threshold. Then, select [Threshold Monitoring] [Set Threshold] menu.
  • Page 53 Chapter 3 Operations of Performance Monitoring Function Figure 3-15 Set Threshold Dialog (When displayed by selecting a cell) Figure 3-16 Set Threshold Menu (Select a line) I-47...
  • Page 54 Chapter 3 Operations of Performance Monitoring Function Figure 3-17 Set Threshold Dialog (When displayed by selecting a line) When you specify a threshold value in the [Set Threshold] area and click the [OK] button, the value is set as a threshold and the dialog box closes.
  • Page 55 Chapter 3 Operations of Performance Monitoring Function (Note 1) A performance monitor server using older version software does not support some of the threshold value setting functions. In such a case, the applicable fields for setting threshold values are displayed in gray, accepting no settings.
  • Page 56 Chapter 3 Operations of Performance Monitoring Function (Note 1) A performance monitor server using older version software may not support some of the threshold monitoring conditions specifying functions. In such a case, the Customize Threshold Monitoring Conditions dialog is not displayed.
  • Page 57 Chapter 3 Operations of Performance Monitoring Function (Note 1) A performance monitor server using older version software may not support some of the threshold excess release specifying functions. In such a case, the Customize Resetting Threshold Excess State dialog is not displayed. For specifying release of threshold excess, software of the latest version is required.
  • Page 58 Chapter 3 Operations of Performance Monitoring Function (6) All thresholds deletion The threshold values for all monitoring targets included in specified disk array subsystem can be deleted by batch processing. On the Numeric Table screen, selecting [Threshold Monitoring] [Delete all Thresholds] and clicking [OK] button on the Delete all Thresholds dialog executes delete all Thresholds.
  • Page 59 Chapter 3 Operations of Performance Monitoring Function Port Displays the name of ports, by disk array, for which a threshold is specified for each monitored item, such as I/O Density, Transfer Rate, Average Transfer Length, and Average Response Time. The port that has a value which has exceeded the threshold is displayed along with the “**”...
  • Page 60 Chapter 3 Operations of Performance Monitoring Function For analyzing statistic information, specify settings to accumulate statistic information history files on the iSM Management server first. By summarizing, extracting, and editing the accumulated statistic information history files on the iSM Management server, the statistic information history is managed. Otherwise, by downloading the accumulated statistic information files and summarizing, extracting, and editing the data on the hard disk of an iSM client, the statistic information history is managed.
  • Page 61 Chapter 3 Operations of Performance Monitoring Function (ii) Stop storing of statistic information Two methods are available to terminate the accumulation of statistic information: a method for specifying a specific disk array and a method for specifying all the disk arrays monitored by the iSM server simultaneously. To terminate the accumulation of statistic information of a specific disk array, assign a disk array name to the end of -d option and then execute the iSMprflog terminate command.
  • Page 62 (3) Setting in the environment definition file The operation of accumulating statistic information can be controlled by performing setting in the environment definition file. For detailed setting procedures, refer to 1.3 “Environment Setting” of the “NEC Storage Manager User’s Manual”.
  • Page 63 Select statistic information files to download, from the [Statistic Information files]. (iii) Specify a downloading target directory for [Download place] field. Specifying by clicking [Browse] is allowed also. By default, data is downloaded in C:\Program Files\NEC\iSMclient\DATA. (iv) Click [Download] button.
  • Page 64 Chapter 3 Operations of Performance Monitoring Function Operations with the Archiver (iSMprfarc) command for summarizing statistic information will be described below. (1) Files to be summarized and determination of the summarizing method Before executing Archiver, specify the following items: Files to be summarized Specify statistic information files you want to summarize.
  • Page 65 Chapter 3 Operations of Performance Monitoring Function iSMprfarc /opt/iSMsvr/etc/mon/20020401NECStorage4100.psl As a result, the primary summarized file (20020401NECStorage4100.psa) on April 1, 2002 is created in the current directory. Example: Example 1 of summarizing when a period is specified Assume that the statistic information history files (20020401NECStorage4100.psl - 20020430NECStorage4100.psl) from April 1, 2002 to April 30, 2002 regarding “NECStorage4100”...
  • Page 66 Chapter 3 Operations of Performance Monitoring Function statistic information files (For details, refer to Part I “Performance Monitoring Function”, 1.5.3.1 “Accumulating Statistic Information” of this manual). [Naming Convention for Statistic Information Files] <YYYY><MM><DD><NN-N>.psl (History file) <YYYY><MM><DD><NN-N>.psa (Primary summarized file) <YYYY><MM><NN-N>.psa (Secondary summarized file) <YYYY>...
  • Page 67 Chapter 3 Operations of Performance Monitoring Function Figure 3-24 Performance Analysis Supporting Tool Screen (Archiver) File Specification Select this function when you want to perform a summarizing operation by specifying statistic information files. This function cannot be selected together with Period Specification. (ii) Statistic Info File Specify statistic information history /summarized files.
  • Page 68 20020430NECStorage4100.psl) from April 1, 2002 to April 30, 2002 regarding “NECStorage4100” are located in the “C:\Program files\NEC\iSMsvr\etc\mon” directory. Among all files in this directory, if you want to summarize the statistic information history files from April 1, 2002 to April 15, 2002 and output primary summarized files to the same directory, enter the following command in a directory and execute it.
  • Page 69 If the current time is April 15, 2002, the statistic information files in the period from April 1, 2001 to March 15, 2002 are the targets of summarizing. iSMprfarc -p “C:\Program files\NEC\iSMsvr\etc\mon” -day 1m As a result, the secondary summarized files (200104NECStorage4100.psa - 200203NECStorage4100.psa) from April, 2001 to March, 2002 are created in the “C:\Program...
  • Page 70 Chapter 3 Operations of Performance Monitoring Function iSMprfext -file 20020401NECStorage4100.psl 20020402NECStorage4100.psa Example 3: Extracting statistic information from the statistic information history file “20020401NECStorage4100.psl” and creating a CSV file in “/temp” iSMprfext -file 20020401NECStorage4100.psl -p /temp Example 4: Extracting statistic information from the statistic information history files “20020401NECStorage4100.psl”...
  • Page 71 Chapter 3 Operations of Performance Monitoring Function Extracting statistic information of the specific type Example 1: Extracting statistic information regarding the number of inputs/outputs iSMprfext -file file1 -iocnt Example 2: Extracting statistic information regarding the number of inputs/outputs and Transfer Rate iSMprfext -file file1 -iocnt -rate Extracting statistic information of the specific component in the specified period Example 1:...
  • Page 72 Chapter 3 Operations of Performance Monitoring Function Specify multiple files to the CSV Conversion Tool and create corresponding CSV files. Link the CSV files created in procedure (a) by using the cat command to put the files together in one CSV file.
  • Page 73 Chapter 3 Operations of Performance Monitoring Function Figure 3-25 Performance Analysis Supporting Tool Screen (CSV Conversion Tool) Statistic Info File Specify statistic information history/summarized files. Multiple files can be specified, and statistic information history files and statistic information summarized files can be mixed. (ii) Output Directory Specify a directory in which output files (CSV files) are created.
  • Page 74 Chapter 3 Operations of Performance Monitoring Function and end time, or either one of them. If no period is specified, statistic information in all periods is extracted. (vii) Output Component Info Select this function to output component information of the disk array in addition to statistic information. (viii) Output Header Info Select this function to output header record showing field name to the head of file.
  • Page 75 Chapter 3 Operations of Performance Monitoring Function iSMprfext -file 20020401NECStorage4100.psl 20020402NECStorage4100.psl -p C:\temp (ii) Extracting specific statistic information You can extract specific statistic information by specifying extracting conditions, such as the period of extraction, component of the disk array, type of statistic information, etc. in the CSV conversion command. Multiple conditions can be specified simultaneously.
  • Page 76 Chapter 3 Operations of Performance Monitoring Function iSMprfext -file file1 -from 20020401/1000 -to 20020401/1530 -ld 20 (e) Extracting statistic information of the specific type in the specified period Example 1: Extracting statistic information regarding the response time after 17:30 on April 1, 2002 iSMprfext -file file1 -from 20020401/1730 -resp (f) Extracting statistic information of the specific component regarding the specific type in the specified period Example 1: Extracting statistic information of the logical disk of 0020h regarding the number of inputs/outputs...
  • Page 77 Chapter 3 Operations of Performance Monitoring Function iSMprfext -file 20020401NECStorage4100.psl 20020402NECStorage4100.psl copy 20020401NECStorage4100_psl.csv+20020402NECStorage4100_psl.csv NECStorage4100.csv This section will explain how to operate the editing tool and the graph making tool. (1) Editing tool You can perform an editing operation by specifying an option in the iSMprfedit command and executing it. Multiple statistic information history/summarized files, output directory, and the type of editing reports to be created can be specified.
  • Page 78 Chapter 3 Operations of Performance Monitoring Function This section will explain how to operate the editing tool and the graph making tool, and the procedures for editing statistic information using spreadsheet software. (1) Editing Tool The editing tool provides the following two operation methods: Operations on the Performance Analysis Supporting Tool screen Select [Performance Analysis Supporting Tool] on [Start] menu to display the Performance Analysis Supporting Tool screen.
  • Page 79 Chapter 3 Operations of Performance Monitoring Function (iii) Select Edit Report Select this function to output editing reports. (iv) Summary Report Specify this function to output summary reports. Detail Report Select the detailed report to be output. (vi) Detail Report About Disk Array Select this parameter to output a detailed report on the whole disk array.
  • Page 80 Chapter 3 Operations of Performance Monitoring Function Example 4: Creating only an detailed report on the whole disk array of the statistic information history file “20020430NECStorage4100.psl”. iSMprfedit -nosum -det 20020430NECStorage4100.psl (2) Graph Making Tool To create graphs, select Storage Performance Report Viewer from the Start menu and specify necessary items on the Storage Performance Report Viewer screen.
  • Page 81 Chapter 3 Operations of Performance Monitoring Function Average Response Time This is the average response time of a single input/output request. The unit is ms (millisecond). Read Hit Ratio This is a ratio of the number of inputs that hit the cache to the total number of input requests. Prefetch Hit Ratio This is a ratio of the number of inputs that hit the cache via the disk array prefetch function to the total number of inputs that hit the cache.
  • Page 82 Chapter 3 Operations of Performance Monitoring Function In “Figure 3-28 Example of Creating a Graph”, some points are not linked by lines. The graph making tool (Storage Performance Report Viewer) does not link points with lines when statistic information is no longer continuous in order to accurately express the continuity of the statistic information in the graph.
  • Page 84: Chapter 4 Overview Of Performance Optimization Function

    The Performance Optimization function (NEC Storage PerformanceOptimizer) is capable of analyzing the causes of lowered responses in the NEC Storage series Disk Array Subsystem and improving its performances. An administrator, with use of the Performance Optimization function, analyzes the performances of disk array under the control of iSM, by centralized operations on an iSM client and tunes the performances as necessary.
  • Page 85: Nec Storage Performanceoptimizer And Related Product

    The NEC Storage PerformanceOptimizer is for the following program products: NEC Storage BaseProduct NEC Storage PerformanceMonitor NEC Storage ReallocationControl NEC Storage AccessControl For system configuration examples, refer to Chapter 3 “Basic Functions” in the “NEC Storage Manager User’s Manual”. II-3...
  • Page 86 Operation flow Reference Installing Performance optimization function Installation of NEC Storage Manager is accompanied by installing Chapter 5 Installation of Performance the Performance optimization function at the same time. For using Optimization Function the Performance optimization function, Work Disk for Optimization must be specified.
  • Page 87: Function Overview

    The NEC Storage PerformanceOptimizer moves logical disks located in a high load physical disk in which I/O concentrates to a low load RANK so as to distribute the concentration of I/O thereby canceling the I/O competition and removing the bottleneck.
  • Page 88 Chapter 4 Overview of Performance Optimization Function Optimization). By moving a logical disk, the arrangement of the Moved Material Logical Disk and the Work Disk for Optimization is switched. The logical disks are moved inside the disk array without affecting any business system. As a result, performance tuning is possible while business is in operation.
  • Page 89: Chapter 5 Installation Of Performance Optimization Function

    The operating environment of the target HW, operation OS, memory, and disk Capacity is the same as that of iSM. For the operating environment, refer to Part I “Installation and Setting of NEC Storage Manager” of the “NEC Storage Manager User’s Manual” of your OS.
  • Page 90: Client Installation

    To install the NEC Storage PerformanceOptimizer, specify a setup type that includes the expanding function and install the iSM client. For the installation method, refer to Part I “Installation and Setting of NEC Storage Manager” of the “NEC Storage Manager User’s Manual” of your OS.
  • Page 91: Installation Of Work Disks For Optimization

    Generally, a bottleneck in performance occurs due to an increase of the amount of data, which results from an increase of the amount of work, or an access pattern change. Work Disks for Optimization are to be used when NEC Storage PerformanceOptimizer eliminates the cause of a bottleneck in performance.
  • Page 92 Chapter 5 Installation of Performance Optimization Function Capacity of a Work Disk for Optimization The capacity of the Moved Material Logical Disk must be identical to that of the Work Disk for Optimization. Set the Work Disk for Optimization in accordance with the capacity of the Moved Material Logical Disk. RAID type of a Work Disk for Optimization A logical disk can be moved even if the RAID type is different between the logical disk and Work Disk for Optimization.
  • Page 93: Chapter 6 Operation Of Performance Optimization Function (Gui

    Chapter 6 Operation of Performance Optimization Function (GUI) This chapter will describe the composition of the NEC Storage PerformanceOptimizer dedicated client screen and its method of operation. Figure 6-1 shows the NEC Storage PerformanceOptimizer dedicated client screen. This section will describe the composition of the screen.
  • Page 94 (4) Configuration Display Area Configuration of the disk array that uses the NEC Storage PerformanceOptimizer is separated into the “disk array layer” and the “RANK layer” and displayed in a tree view form. Figure 6-2 shows an example of the display.
  • Page 95 Chapter 6 Operation of Performance Optimization Function (GUI) Table 6-1 Disk Array Layer and RANK Layer Layer Description Disk Array layer Displays icons and the disk array name which correspond to respective disk array models. An icon for the disk array which is not monitored by the iSM server is displayed in gray. RANK layer Displays the number and icon for the RANKs built in the disk array.
  • Page 96 Chapter 6 Operation of Performance Optimization Function (GUI) Table 6-2 Detailed Information Display Items for the Disk Array Display item Description Disk Array Displays the disk array name. Subsystem Monitoring State Running ..Indicates that the iSM server is monitoring the state. Stop ..Indicates that the iSM server is stopping monitoring the state.
  • Page 97 Chapter 6 Operation of Performance Optimization Function (GUI) Figure 6-5 Example of Display of Detailed Information of Logical Disk Table 6-4 Detailed Information Display Items for the Logical Disk Display item Description Number Displays icons and numbers of logical disks. Similar to RANKs’ icons, logical disks’ icons are also displayed in the three stages expressed in different colors depending on Busy Ratio.
  • Page 98 Chapter 6 Operation of Performance Optimization Function (GUI) The optimization logical disk is bound for a different RANK from the logical disk selected in the detailed display area. Figure 6-6 shows an example of display of detailed information of the Work Disk for Optimization. Table 6-5 provides descriptions of each display item.
  • Page 99 Chapter 6 Operation of Performance Optimization Function (GUI) Table 6-6 Detailed Information Display Items for Movement Progress Information Display item Description Disk Array Subsystem Displays the disk array name for a logical disk being moved. Source RANK Displays the RANK number for a logical disk being moved. Source LD Displays the logical disk number being moved.
  • Page 100 Chapter 6 Operation of Performance Optimization Function (GUI) (8) Status Bar Displays the operating state of the NEC Storage PerformanceOptimizer client. Table 6-7 provides descriptions of each display item. Table 6-7 Display Items for Status Bar Display item Description Operating state Displays the current operating state.
  • Page 101: Operation On Screen

    Chapter 6 Operation of Performance Optimization Function (GUI) The Busy Ratios of RANKs and logical disks are displayed on the screen. The displayed values are average values during the specified period. This period is called an analysis period. The time zone in which the disk array is used depends on the business system. The following is the examples. The time zone in which the disk array is used is limited to a specific time zone for a day.
  • Page 102 Chapter 6 Operation of Performance Optimization Function (GUI) Period specification area Days of the week specification area Time zone specification area Figure 6-8 Example of Display of Analysis Property Dialog ([Analysis Period] tab) Table 6-8 Descriptions of the areas in Analysis Property dialog ([Analysis Period] tab) Area Description Period specification area...
  • Page 103 Chapter 6 Operation of Performance Optimization Function (GUI) Select this function to analyze a specific period in the past. Period to calculate Select this function to analyze data in a specified period periodically such as daily, weekly or monthly. For example, if you set the period from six days ago to the present, you can perform weekly analysis of the week without changing settings.
  • Page 104 Chapter 6 Operation of Performance Optimization Function (GUI) procedures. (1) Displaying Analysis Property dialog (“Sort of Busy Ratio” tab) Select [Setting] [Analysis Property] menu on the main screen of the performance optimization function. On the displayed Analysis Property dialog box, select [Sort of Busy Ratio] tab that displays an area for specifying [Sort of Busy Ratio].
  • Page 105 Chapter 6 Operation of Performance Optimization Function (GUI) (2) Procedures for setting Sort of Busy Ratio This section describes the procedures for specifying the Sort of Busy Ratio on the Analysis property dialog ([Sort of Busy Ratio] tab). The Sort of Busy Ratio can be saved by selecting [File] [Save Settings] menu on the main screen of the performance optimization function, for settings upon activating next time.
  • Page 106 Chapter 6 Operation of Performance Optimization Function (GUI) Figure 6-10 Analysis Property Dialog ([Folder to Store Data] tab) Table 6-10 Description of Analysis Property dialog (“Folder to store data” tab) Area Description Folder to Store Data Specifies a folder to be a folder to store data. (2) Setting Folder to Store Data Enter an arbitrary folder as a folder to store data, or click the [Browse] button and select a directory on the Browse for Folder dialog.
  • Page 107 Chapter 6 Operation of Performance Optimization Function (GUI) by one of the following three procedures: Popup menu Select disk array or RANK on the Configuration Display Area or Detailed Display Area, and display a popup menu by right-clicking. Select [Busy Ratio Graph] on the popup menu. Menu Select disk array or RANK on the Configuration Display Area or Detailed Display Area, and select [Operation] [Busy Ratio Graph].
  • Page 108 Chapter 6 Operation of Performance Optimization Function (GUI) LD selection area (when RANK is selected on Configuration display area or Detailed Display Area), or RANK selection area (when Disk Array Subsystem is selected on Configuration display area or Detailed Display Area) Target period area Figure 6-11 [Busy Ratio Graph] Dialog Table 6-11 Descriptions of The Busy Ratio Graph Dialog...
  • Page 109 Chapter 6 Operation of Performance Optimization Function (GUI) Figure 6-12 Download Selection Dialog Table 6-12 [Download Selection] dialog Area Description Downloads optimization log file recording Busy Show graph after downloading insufficient data Ratios onto the iSM client. Displays Busy Ratio Graph without downloading Show graph using data already downloaded optimization log file.
  • Page 110 Chapter 6 Operation of Performance Optimization Function (GUI) iSM client are updated and the data older than one month old is deleted. Upon attempting execution of download process, if a performance optimization log file storing Busy Ratios does not exist on the iSM server, the process results in an error without executing downloading. Meanwhile, upon attempting execution of Busy Ratio Graph without downloading performance optimization log file, the process results in an error without showing the Busy Ratio Graph if no data exists in the folder to store data of the iSM client.
  • Page 111 Chapter 6 Operation of Performance Optimization Function (GUI) Table 6-13 Description of The Busy Ratio Graph screen Area Description Selects a time range of Busy Ratio Graphed on screen. A range is specified by hours (3/6/12/24), days (1 to 31), or weeks (1 to 4). Setting is allowed within a Graph scale setting area period specified by the target period area on the Busy Ratio Graph dialog.
  • Page 112 Chapter 6 Operation of Performance Optimization Function (GUI) This function detects High Load RANK and makes the most effective replacement plan. A list of replacement plans expecting higher effects of I/O distribution after moving is made for replacing source/destination logical disks. The replacement plan is applicable to the RANKs reaching danger zone and critical zone specified on the Analysis Property dialog (“Sort of Busy Ratio”...
  • Page 113 Chapter 6 Operation of Performance Optimization Function (GUI) Table 6-14 Description of [Replacement Plan] Dialog Area Description Disk Array Subsystem Nicknames of selected disk arrays are displayed. RANK numbers of the bottleneck RANKs area displayed. The applicable RANKs to be displayed in the replacement plan display area can be changed by High Load RANK selecting RANKs on a pull-down menu.
  • Page 114 Chapter 6 Operation of Performance Optimization Function (GUI) The performance optimization function is capable of estimating an effect of load distribution after logical disk movement. By executing this function, the changes of Busy Ratios in source and destination logical disk or RANKs are displayed in graphs.
  • Page 115 Chapter 6 Operation of Performance Optimization Function (GUI) Table 6-15 The [Graph Indicating Replacement Effect Prediction] dialog Area Description Specifies start date and end date of a period for displaying Replacement Target period area Effect Prediction graph. Specify period for Replacement Effect Prediction On the [Graph Indicating Replacement Effect Prediction] dialog, specify a period for Replacement Effect Prediction of logical disks.
  • Page 116 Chapter 6 Operation of Performance Optimization Function (GUI) Figure 6-18 Confirmation of Overwriting Dialog Clicking [OK] on the Confirmation of Overwriting dialog overwrites the existing data. Click the [Change Folder] button allows changing the folders to store data to be downloading destination. For changing the folder to store data, refer to 6.2.3 “Setting Folder to Store Data”.
  • Page 117 Chapter 6 Operation of Performance Optimization Function (GUI) Graph display area Graph scale setting area Legend display area Figure 6-19 Replacement Effect Prediction Screen Table 6-17 Description of Replacement Effect Prediction screen (1/2) Area Description Selects a time range of Replacement Effect Prediction graph displayed on one screen.
  • Page 118 You can specify a RANK in which the Busy Ratio has not reached the critical zone or attention zone yet as a tuning target. It is possible to perform tuning for precautions or perform a tuning operation when short-term I/O concentration has been detected by NEC Storage PerformanceMonitor. RANK to be tuned Figure 6-20 Selecting a RANK to Be Tuned Figure 6-20 shows an example of selecting a RANK to be tuned in the detailed information display area.
  • Page 119 Chapter 6 Operation of Performance Optimization Function (GUI) Double-clicking the target RANK will display logical disks bound for the RANK in the Detailed Display Area. (ii) Selecting Moved Material Logical Disks/Selecting Work Disk for Optimization Among logical disks bound for the RANK to be tuned, select a Moved Material Logical Disk. It is not always right to move a logical disk which has the highest Busy Ratio.
  • Page 120 Chapter 6 Operation of Performance Optimization Function (GUI) The Busy Ratio of the source RANK is displayed in the Double-clicking a logical disk “Before movement After will display destination candidate movement” format. Work Disks for Optimization. The Busy Ratio of the destination RANK is displayed in the “Before movement After...
  • Page 121 Chapter 6 Operation of Performance Optimization Function (GUI) Movement Speed area Start Schedule area Figure 6-22 Start LD Movement Dialog Table 6-18 Description of Start LD Movement Dialog Area Description Select Moving speed between “auto” and “high”. The following describes the two speeds.
  • Page 122 Chapter 6 Operation of Performance Optimization Function (GUI) When registering a LD movement schedule of the logical disks, specify a movement speed and select [Start at Specified Time] and specify moving start time and time of continuing re-execution. Click the [OK] button. A confirmation dialog appears.
  • Page 123 After completion of logical disk movement, the relation between the logical disk and the RANK changes. This configuration change is detected by the iSM simultaneously with the completion of the transfer and is reflected on the state monitoring screen of the iSM as well as on the NEC Storage PerformanceOptimizer screen. (3) Exceptional Operations After the transfer of a logical disk has started or the movement schedule has registered, there is a possibility of canceling the movement, changing the speed and changing (canceling) the movement schedule for some reason.
  • Page 124 Chapter 6 Operation of Performance Optimization Function (GUI) Tool Bar button Click the [Stop LD Movement] button on the tool bar. Figure 6-24 shows an example of Stop LD Movement dialog. When you want to cancel transfer, click the [OK] button on the Stop LD Movement dialog screen. Figure 6-24 Stop LD Movement Dialog (iii) Cancellation When you click the [OK] button on the Stop LD Movement dialog screen, the confirmation dialog...
  • Page 125 Chapter 6 Operation of Performance Optimization Function (GUI) Figure 6-25 Change LD Movement Speed Dialog (iii) Specifying movement speed after change Select auto or high in the movement speed change area. (iv) Executing change of moving speed Clicking [OK] button on the Change LD Movement Speed dialog displays a confirmation dialog. Clicking [Yes] on the confirmation dialog changes the moving speeds.
  • Page 126 Chapter 6 Operation of Performance Optimization Function (GUI) Figure 6-26 Change LD Movement Schedule Dialog (iii) Change (cancel) movement schedule For changing time of starting disk movement, or time of continuing re-execution Change the setting value and click the [Change Schedule] button. For canceling disk movement schedule Click the [Cancel Schedule] button.
  • Page 127: Chapter 7 Operation Of Performance Optimization Function (Cli

    Function (CLI) This chapter explains how to operate a command provided by NEC Storage PerformanceOptimizer. The command provided by NEC Storage PerformanceOptimizer can be executed on the server on which NEC Storage Manager is operating. As explained previously, logical disks can be moved through operation on the client screen of NEC Storage PerformanceOptimizer.
  • Page 128 Message for the iSMprfctl command: iSMprfctl 001 Message output to the iSM operation log: iSM09002 Completion of movement Make sure of the NEC Storage PerformanceOptimizer message that is output to the iSM operation log. Message output to the iSM operation log: iSM09003 II-46...
  • Page 129 The iSMprfctl command is executable only on the server on which NEC Storage Manager is operating. Do not execute the command on a Business Server or while NEC Storage Manager is in stopped state. The iSMprfctl command is available with a client PC only when Telnet is connected. It is not available with Terminal Service.
  • Page 130: Appendix A Commands Of Performance Monitoring/Optimization Function

    Appendix A Commands of Performance Monitoring/Optimization Function Appendix A Commands of Performance Monitoring/Optimization Function [ NAME ] iSMprflog - Operates accumulation of statistic information log file. [ SYNOPSIS ] iSMprflog start [-a] [-d Disk Array Subsystem name] iSMprflog stop [-a] [-d Disk Array Subsystem name] [ DESCRIPTI/ON ] iSMprflog command starts/stops accumulation of statistic information log file.
  • Page 131 Appendix A Commands of Performance Monitoring/Optimization Function [ USAGE ] Example: Start accumulation of statistic information log file of all disk array monitored by the iSM server. iSMprflog start -a Example: Start accumulation of statistic information log file of a disk array (NECStorage4000/10) monitored by the iSM server.
  • Page 132 Appendix A Commands of Performance Monitoring/Optimization Function current directory. The names of the output files follow the naming convention for statistic information files (See Remark). A file name indicates the disk array name from which statistic information is collected and the date (or month) of collection. If the file with the same name exists, the file is overwritten.
  • Page 133 Appendix A Commands of Performance Monitoring/Optimization Function iSMprfarc selects the statistic information files corresponding to the specified period from the files in the specified directory, and summarizes statistic information in the specified unit of summarizing. If statistic information files for multiple disk arrays exist in the directory, it performs summarizing process for each disk array.
  • Page 134 Appendix A Commands of Performance Monitoring/Optimization Function -hour 30d Example: The target period of secondary summarizing or direct secondary summarizing is 2 months before the present date and earlier. (If the present date is August 15, 2002, the target period of summarizing is before February 15, 2002.) -day 2m The -v option outputs the detailed information of summarizing process in the standard output.
  • Page 135 Appendix A Commands of Performance Monitoring/Optimization Function The value is provided as the average response time after summarizing. More specifically, in each interval of the input file, the product of each interval time and each average response time is calculated. The products are summed within the period of summarizing, and the total response time in the period of summarizing is calculated.
  • Page 136 Appendix A Commands of Performance Monitoring/Optimization Function period of summarizing are provided as additional information. If the input file is a statistic information history file (primary summarizing and direct secondary summarizing), I/O count in each logging Interval is divided by each logging Interval time, and time density of I/O count is calculated. Then, the maximum and minimum values in the period of summarizing are selected.
  • Page 137 Appendix A Commands of Performance Monitoring/Optimization Function [ NOTES ] Naming Convention for Statistic Information Files The year, month, and day in the file name indicates the day or month of collection of the statistic information. YYYYMMDDNNN~NN.psl : Statistic information history file (in n minutes/day; n depends on the configuration) YYYYMMDDNNN~NN.psa : Primary summarized file (every hour for one day)
  • Page 138 Appendix A Commands of Performance Monitoring/Optimization Function [ NAME ] iSMprfext- Extracts statistic information, converts it to the CSV format, and outputs it to a file. [ SYNOPSIS ] iSMprfext -file File Name [File Name..] [ -from Time ] [ -to Time ] [ -hp [ Host Port [Host Port..
  • Page 139 Appendix A Commands of Performance Monitoring/Optimization Function Month [01-12] Day (in one month)[01-31] Hour (in 24-hour time)[00-23] Minute [00-59] -to Time Specify the end time of extraction to extract statistic information by specifying a period. The method and interpretation of time are the same as the -from option. -hp Host Port (Port in Host Director) Specified if you want to extract statistic information of the specified host port.
  • Page 140 Appendix A Commands of Performance Monitoring/Optimization Function -pd Physical Disk Specified if you want to extract statistic information of the specified physical disk. For the identification number, specify the physical disk number or the range of logical disk numbers. Up to 10 physical disks can be specified.
  • Page 141 Appendix A Commands of Performance Monitoring/Optimization Function Average Response Time Maximum Response Time -rate Specified if you want to extract statistic information regarding the Transfer Rate. Statistic information regarding the Transfer Rate is the following. Transfer Rate Maximum Transfer Rate Minimum Transfer Rate -busy Specified if you want to retrieve statistic information regarding a busy ratio.
  • Page 142 Appendix A Commands of Performance Monitoring/Optimization Function The following extracts statistic information from 10:30 to 14:00 on April 30, 2002. iSMprfext -file file1 -from 20020430/1030 -to 20020430/1400 The following extracts statistic information of host ports with the director number: 10 and port number: 1e.
  • Page 143 Appendix A Commands of Performance Monitoring/Optimization Function The following extracts only statistic information of disk ports with the director number: 30 and port number: 10 as well as physical disks with the group number: 2 and disk number: a. iSMprfext -file file1 -dp 30-10 -pd 2-a The following extracts only statistic information regarding the number of inputs/outputs.
  • Page 144 Appendix A Commands of Performance Monitoring/Optimization Function [ SYNOPSIS ] iSMprfedit [ -o directory name] [ -nosum ] [ -det [ [ -all ] | [ -array ] [ -rank ] [ -hp ] [ -dp ] [ -ld ] [ -pd ] ] ] File Name [File Name..
  • Page 145 Appendix A Commands of Performance Monitoring/Optimization Function xxxxxx_psa_edit type.txt In the above “edit type” is specified one of the following strings depending on the type of report: “sum” : Performance summary report “array” : Entire disk array performance detail report “rank”...
  • Page 146 Appendix A Commands of Performance Monitoring/Optimization Function Outputs a physical disk performance detail report. File Name Specify the statistic information history file (.psl) and statistic information summarized file (.psa) to be edited. [ DIAGNOSTICS ] The iSMprfedit command, on normal termination, outputs a normal termination message in the standard output, and in abnormal termination, outputs an error message in the standard error output.
  • Page 147 Appendix A Commands of Performance Monitoring/Optimization Function for Windows) iSMprfedit -o C:\tmp file1 file2...
  • Page 148: Command For Performance Tuning

    Appendix A Commands of Performance Monitoring/Optimization Function [ NAME ] iSMprfctl - Moves logical disks. [ SYNOPSIS ] iSMprfctl start_ldmv -arrayname Disk Array Subsystem name | -ldn logical-disk-number | -ldname [volume-format:] logical-disk-name | -wkldn logical-disk-number | -wkldname [volume-format:] logical-disk-name [ -speed | auto | high | ] [ DESCRIPTI/ON ] The iSMprfctl command moves logical disks for optimizing performance.
  • Page 149 Appendix A Commands of Performance Monitoring/Optimization Function -ldn Logical disk number of the move-source logical disk Specify the logical disk number of the logical disk to be moved. If the logical disk number is specified with “h” suffixed, it is processed as a hexadecimal number. With “h” being omitted, it is processed as a decimal number.
  • Page 150 Appendix A Commands of Performance Monitoring/Optimization Function [USAGE ] Example: Move data between the logical disk 000ah and the work disk for optimization 0010h. Select the default LD movement speed (auto). iSMprfctl start_ldmv -arrayname NECStorage4100/10 -ldn ah -wkldn 10h Example: Move data between the logical disk 10 and the work disk for optimization 16. Select the default LD movement speed (auto).
  • Page 151: Appendix B Csv File Structure

    Appendix B CSV File Structure Appendix B CSV File Structure A CSV file is a set of records with data separated by commas. Records are separated by a line feed character. The maximum number of records is 65,536. The types of records which comprise a CSV file are as follows.
  • Page 152 Appendix B CSV File Structure (2) Statistic information record This record contains statistic information for each component of the disk array. There are the following types of statistic information records. Statistic information record of a host port (a port of the host director) Statistic information record of a disk port (a port of the disk director) Statistic information record of a physical disk Statistic information record of a logical disk...
  • Page 153 Appendix B CSV File Structure Port No Port number ‘Hexadecimal’ Port number of the host port Write Count Write I/O count Decimal Total count of Write I/O in period of accumulation/summarizing. Average Write Average response Decimal Average response time (microsecond) of Response Time time of Write I/O Write I/O in period of...
  • Page 154 Appendix B CSV File Structure Max Read Max. transfer rate Decimal Maximum transfer rate (kilobyte/second) of Read I/O Transfer Rate of Read I/O in each accumulation period (KB/S) in the period of summarizing. Min Read Min. transfer rate Decimal Minimum transfer rate (kilobyte/second) Transfer of Read I/O of Read I/O in each accumulation period...
  • Page 155 Appendix B CSV File Structure Max Write Max. of I/O Decimal Maximum value of I/O density of Write Count density of Write I/O in each accumulation period in the period of summarizing. Min Write Min. of I/O Decimal Minimum value of I/O density of Write Count density of Write I/O in each accumulation period in the...
  • Page 156 Appendix B CSV File Structure Average Write Average response Decimal Average response time (microsecond) of time of Write I/O Response Time Write I/O in period of (micro sec) accumulation/summarizing Max Write Maximum Decimal Maximum response time (microsecond) Response Time response time of of Write I/O in period of (micro sec) Write I/O...
  • Page 157 Appendix B CSV File Structure Min Busy Min. of average Decimal Minimum value of the average busy Ratio[%] busy ratio ratio in each accumulation period in the period of summarizing (iv) Statistic information record format for logical disks Save Field name Location (-header specified) Content...
  • Page 158 Appendix B CSV File Structure Average Read Average response Decimal Average response time (microsecond) of time of Read I/O Response Time Read I/O in period of (micro sec) accumulation/summarizing Max Read Maximum Decimal Maximum response time (microsecond) Response Time response time of of Read I/O in period of (micro sec) Read I/O...
  • Page 159 Appendix B CSV File Structure Min Write Hit Min. of I/O Decimal Minimum value of I/O density of Count density of CacheHit Write I/O in each CacheHit Write accumulation period in the period of summarizing. Max Write Hit Max. transfer rate Decimal Maximum transfer rate (kilobyte/second) Transfer...
  • Page 160 Appendix B CSV File Structure Max Busy Max. of average Decimal Maximum value of the average busy Ratio[%] busy ratio ratio in each accumulation period in the period of summarizing Min Busy Min. of average Decimal Minimum value of the average busy Ratio[%] busy ratio ratio in each accumulation period in the...
  • Page 161 Appendix B CSV File Structure Read Transfer Transfer rate of Decimal Transfer rate (kilobyte/second) of Read Rate(KB/S) Read I/O I/O in period of accumulation/summarizing Busy Ratio[%] Average busy Decimal Average busy ratio (percent) in the ratio period of accumulation/summarizing Busy Time Operating Time Decimal Operating Time (milli sec) in the period...
  • Page 162 Appendix B CSV File Structure “Save location” in the table indicates the order of the data separated by commas in the record. The save location of the leftmost data is 1. Configuration information record format for disk array Save Field name Location (-header specified) Content...
  • Page 163 Appendix B CSV File Structure (iii) Configuration information record format for disk ports Save Field name Location (-header specified) Content Format Description R_COMPO_DP Record identifier String “R_COMPO_DP” is set. Date Date yyyy/mm/dd Date when configuration information in the record was collected. yyyy-Year mm-Month dd-Day...
  • Page 164 Appendix B CSV File Structure Configuration information record format for physical disks Save Field name Location (-header specified) Content Format Description R_COMPO_PD Record identifier String “R_COMPO_PD” is set. Date Date yyyy/mm/dd Date when configuration information in the record was collected. yyyy-Year mm-Month dd-Day...
  • Page 165 Appendix B CSV File Structure (vii) LD Name Change Record Format Save Field name Location (-header specified) Content Format Description R_COMPO_NNC Record identifier String “R_COMPO_NNC” is set. Date Date yyyy/mm/dd Date when the logical disk name was changed. yyyy-Year mm-Month dd-Day Time Time...
  • Page 166 Appendix B CSV File Structure (ix) Logical disk move record format Save Field name Location (-header specified) Content Format Description R_COMPO_LDM Record identifier String “R_COMPO_LDM” is set. Date Date yyyy/mm/dd Date when the logical disk has been transferred. yyyy-Year mm-Month dd-Day Time Time...
  • Page 167: Appendix C Report Format Of Performance Report Editor

    Appendix C Report Format of Performance Report Editor Appendix C Report Format of Performance Report Editor The following explains the contents of the statistic information edit report created by the performance report editor (iSMprfedit command). (1) Performance summary report Displays the statistic information on the entire disk array and each component of the disk array throughout the period of summarizing.
  • Page 168 Appendix C Report Format of Performance Report Editor Explanation of display items Display item Explanation disk array name Nickname of the disk array sampling term Period of accumulation of statistic information sampling time unit Unit of accumulation of statistic information “S”: In seconds “H”: in hours “D”: In days...
  • Page 169 Appendix C Report Format of Performance Report Editor Explanation of display items Display item Explanation disk array name Nickname of the disk array Leftmost position (configuration An asterisk is displayed immediately after change in the disk array change mark) configuration or restart of accumulation of statistic information. date, time Date and time statistic information was collected iops...
  • Page 170 Appendix C Report Format of Performance Report Editor Explanation of display items Display item Explanation disk array name Nickname of the disk array port no Number of the host director port Leftmost position (configuration An asterisk is displayed immediately after change in the disk array change mark) configuration or restart of accumulation of statistic information.
  • Page 171 Appendix C Report Format of Performance Report Editor (4) Disk director port performance detail report Chronologically displays port-based statistic information on the disk director port extracted from the statistic information file. Below is shown a display image (file name: 20010529NECStorage4100_psl_dp.txt). <<...
  • Page 172 Appendix C Report Format of Performance Report Editor (5) Logical disk performance detail report Chronologically displays disk-based statistic information on the logical disk extracted from the statistic information file. Below is shown a display image (file name: 20010529NECStorage4100_psl_ld.txt). << LOGICAL DISK PERFORMANCE REPORT >> disk array name NECStorage4100 disk no = 0000h...
  • Page 173 Appendix C Report Format of Performance Report Editor w.ratio Ratio (percent) of write I/O in the accumulation period cache hit Ratio (percent) of cache-hit I/O in the accumulation period read: Ratio (percent) of cache-hit read I/O prefetch: Ratio (percent) of prefetch-hit read I/O among cache-hit read I/O write: Ratio (percent) of cache-hit write I/O busy...
  • Page 174 Appendix C Report Format of Performance Report Editor Explanation of display items Display item Explanation disk array name Nickname of the disk array disk no Number of the physical disk Leftmost position (configuration An asterisk is displayed immediately after change in the disk array change mark) configuration or restart of accumulation of statistic information.
  • Page 175 Appendix C Report Format of Performance Report Editor (7) RANK performance detail report Chronologically displays RANK-based statistic information on the RANK retrieved from the statistic information file. A display image (file name :20020927NECStorage4200_psa_ rank.txt) is shown below. << RANK PERFORMANCE REPORT >> disk array name NECStorage4100 rank no = 00h-00h...
  • Page 176 Appendix C Report Format of Performance Report Editor Explanation of display items Display item Explanation disk array name Nickname of the disk array rank no RANK number First column An asterisk is displayed immediately after changes have been made to the disk array configuration or the accumulation of statistic information has been restarted.
  • Page 177: Appendix D Estimation Of Nec Storage Performancemonitor File Size

    Appendix D Estimation of NEC Storage PerformanceMonitor File Size Appendix D Estimation of NEC Storage PerformanceMonitor File Size The size of a statistic information history file and a statistic information summarized file can be estimated by applying the number of disk array components (the number of host ports (HPs), disk...
  • Page 178 Appendix D Estimation of NEC Storage PerformanceMonitor File Size Furthermore, if there is a period in which there is no input or output during the statistic information storage period, the actual file size may become smaller than the calculated size.
  • Page 179: Appendix E Estimating Size Of Nec Storage Performanceoptimizer Files

    Appendix E Estimating Size of NEC Storage PerformanceOptimizer Files Appendix E Estimating Size of NEC Storage PerformanceOptimizer Files The size of a performance optimizing log file can be estimated by applying the number of disk array components (the number of RANKs and logical disks (LDs)), the storage period of statistic...
  • Page 180 This section presents a case study of building a 10-thousand user scale mail-server system using the NEC Storage series disk devices. The case shows an example of analyzing the causes of failing in achieving performance capabilities, using the performance monitoring function.
  • Page 181 Appendix F Example of Performance Improvement Client’s log-in Client’s log-off Limit of disk capability Mail message sending/receiving by clients is 30000 to 40000 messages/5 minutes Write I/O counts Read I/O counts Total I/O counts Number of disk I/O accesses increases when clients’...
  • Page 182 ............I-67, I-68 entire disk array performance detail report..........I-24, 14, 37 environment setting ........... I-25, II-7 estimating size of NEC Storage PerformanceOptimizer files........48 basic statistic information ......... I-15, I-19 estimation of NEC Storage bottleneck detection support ........II-5 PerformanceMonitor file size ........
  • Page 183 ..............II-38 for statistic information history files......I-14 statistic information history file ........46 NEC Storage AccessControl ........II-3 statistic information primary summarized file ...I-20 NEC Storage Manager..........I-2 statistic information record........20, 21 NEC Storage PerformanceOptimizer....II-2, II-7 statistic information record format NEC Storage ReallocationControl ......
  • Page 184 statistic log storing ............1 status bar..............I-28 stop................I-28 unit to sum up ...............I-9 stop storing of statistic information ......I-51 storing of statistic information ........I-50 summarizing period ............21 summarizing statistic information..... I-19, I-54 support for analysis of statistic information ....I-50 viewing movement progress operation/register movement schedule ..........