Front cover IBM Eserver p5 590 and 595 System Handbook Component-based description of the hardware architecture A guide for machine type 9119 models 590 and 595 Capacity on Demand explained Peter Domberg Nia Kelley TaiJung Kim Ding Wei ibm.com/redbooks...
Page 3
International Technical Support Organization IBM Eserver p5 590 and 595 System Handbook March 2005 SG24-9119-00...
3-10 Link Aggregation (EtherChannel) pseudo device ....78 3-11 IBM p5-590 and p5-595 Virtualization Technologies ....80 3-12 AIX 5L Version 5.3 Virtual I/O Server and client partitions .
Page 12
7-15 Potential system components that require fixes ....187 7-16 Getting fixes from the IBM Sserver Hardware Information Center . . 188 7-17 Partition profile power-on properties ......189 7-18 System Management Services (SMS) main menu .
Page 13
7-21 Current boot sequence menu (default boot list) ....193 Private direct network ........203 HMC with hub/switch attachment.
Page 14
Previous pSeries documentation ......272 IBM Eserver p5 590 and 595 System Handbook...
IBM representative for information on the products and services currently available in your area. Any reference to an IBM product, program, or service is not intended to state or imply that only that IBM product, program, or service may be used. Any functionally equivalent product, program, or service that does not infringe any IBM intellectual property right may be used instead.
You may copy, modify, and distribute these sample programs in any form without payment to IBM for the purposes of developing, using, marketing, or distributing application programs conforming to IBM's application programming interfaces.
Page 20
AIX 5L. He provides clients with technical support on pSeries systems, AIX 5L, and system management. Ding Wei is an Advisory IT Specialist working for IBM China ATS. He has eight years of experience in the Information Technology field. His areas of expertise include pSeries®...
Arsi Kortesnemi (IBM Finland) Become a published author Join us for a two- to six-week residency program! Help write an IBM Redbook dealing with specific products or solutions, while getting hands-on experience with leading-edge technologies. You'll team with IBM technical professionals, Business Partners and/or clients.
Use the online Contact us review redbook form found at: ibm.com/redbooks Send your comments in an email to: redbook@us.ibm.com Mail your comments to: IBM Corporation, International Technical Support Organization Dept. JN9B Building 905 11501 Burnet Road Austin, Texas 78758-3493 IBM Eserver p5 590 and 595 System Handbook...
1.1 Introduction The IBM Sserver p5 590 and IBM Sserver p5 595 are the servers redefining the IT economics of enterprise UNIX and Linux® computing. The up to 64-way p5-595 server is the new flagship of the product line with nearly three times the commercial performance (based on rperf estimates) and twice the capacity of its predecessor, the IBM Sserver pSeries 690.
Page 25
Flexibility in packaging High-density 24-inch system frame for maximum growth. See 1.3, “General overview and characteristics” on page 4 for more information. Shared processor pool Provides the ability to transparently share processing power between partitions. Helps balance processing power and ensures the high priority partitions receive the processor cycles they need.
Optional I/O Drawer or Optional I/O Drawer or Internal batteries Internal batteries First I/O Drawer (required) First I/O Drawer (required) Optional I/O Drawer Optional I/O Drawer Figure 1-1 Primary system frame organization IBM Eserver p5 590 and 595 System Handbook...
Both systems are powered by IBMs most advanced 64-bit Power Architecture microprocessor, the POWER5 microprocessor, with simultaneous multi-threading that makes each processor logically appear as two to the operating system, thus increasing commercial throughput and system utilization over servers without this capability.
LPARs running AIX 5L Version 5.3 or AIX 5L Version 5.2 with the 5200-04 Recommended Maintenance package. IBM Sserver p5 590 and 595 servers also offer optional Capacity on Demand (CoD) capability for processors and memory. CoD functionality is outlined in Chapter 4, “Capacity on Demand”...
Processor speed Better performance ½ proc. speed ½ proc. speed Inter MCM data bus 50% more transistors in Size 412mm 389mm the same space * Based on IBM rPerf projections Figure 1-3 POWER4 and POWER5 architecture comparison Chapter 1. System overview...
These caches allow the p5-590 and p5-595 to run workloads significantly faster than predecessor servers. The difference of memory hierarchy between POWER4 and POWER5 systems is represented in Figure 1-4 as follows: IBM Eserver p5 590 and 595 System Handbook...
Ultra3 SCSI controllers. I/O drawers can be installed in the primary 24-inch frame or in an optional expansion frame. Attachment to a wide range of IBM TotalStorage® storage system offerings – including disk storage subsystems, storage area network (SAN) components, tape libraries, and external media drives –...
Figure 1-5 p5-590 and p5-595 I/O drawer organization 1.3.4 Media bays You can configure your IBM Sserver p5 595 (9119-595) and p5 590 (9119-590) systems to include a storage device media drawer. This media drawer can be mounted in the CEC rack with three available media bays, two in the front and one in the rear.
The POWER5 processor-based Sserver p5 products supports IBM AIX 5L Version 5.2, IBM AIX 5L Version 5.3, SUSE Linux Enterprise Server 9 (SLES 9), and Red Hat Enterprise AS Linux 3 (RHEL AS 3). IBM i5/OS V5R3 is also available on Sserver p5 models 570, 590, and 595.
* 32 GB memory cards to enable maximum memory are planned for availability April 8, 2005. Until that time, maximum memory is half as much (512 GB on p5-590 and 1024 GB on p5-595). IBM Eserver p5 590 and 595 System Handbook...
1.5 Operating systems support All new POWER5 processor-based servers are capable of running IBM AIX 5L Version 5.3 or AIX 5L Version 5.2 for POWER and support appropriate versions of Linux. Both of the aforementioned supported versions of AIX 5L have been specifically developed and enhanced to exploit and support the extensive RAS features on IBM Sserver pSeries systems.
Find full information about SUSE Linux Enterprise Server 9 for POWER at: http://www.suse.com/us/business/products/server/sles/i_pseries.html For the latest in IBM Linux news, subscribe to the Linux Line. See: https://www6.software.ibm.com/reg/linux/linuxline-i Many of the features that are described in this document are OS-dependent and may not be available on Linux.
Page 37
IBM only supports the Linux systems of clients with a SupportLine contract that covers Linux. Otherwise, the Linux distributor should be contacted for support. Chapter 1. System overview...
Page 38
IBM Eserver p5 590 and 595 System Handbook...
2.1 Server overview The IBM Sserver p5 590 and 595 provides an expandable, high-end enterprise solution for managing the computing requirements necessary to become an on demand business. With the introduction of the POWER5 architecture, there has been numerous improvements over the previous POWER4 architecture based systems.
symmetric multi-processor. Going beyond 32 processors with POWER4 architecture could increase interprocessor communication, resulting in higher traffic on the interconnection fabric bus. This can cause greater contention and negatively affect system scalability. Moving the L3 cache reduces traffic on the fabric bus and enables POWER5 processor-based systems to scale to higher levels of symmetric multi-processing.
Adjustable thread priority Adjustable thread priority lets software determine when one thread should have a greater (or lesser) share of execution resources. The POWER5 processor supports eight software-controlled priority levels for each thread. IBM Eserver p5 590 and 595 System Handbook...
Single threaded operation Not all applications benefit from simultaneous multi-threading. Having threads executing on the same processor does not increase the performance of processor intensive applications or applications that consume all of the chip’s memory bandwidth. For this reason, the POWER5 processor supports the single thread (ST) execution mode.
Page 44
SMP scaling requirements dictate a single design to address both environments. POWER5 POWER5 technology is the next generation of IBM 64-bit architecture. Although the hardware is based on POWER4, POWER5 is much more than just an improvement in processor or chip design. It is a major architectural change, creating a much more efficient superscalar processor complex.
Page 45
Memory and CPU sharing, a dual clock, and dual service processors with failover capability are examples of the full system design approach for high availability. IBM designed the Sserver p5 system processor, caching mechanisms, memory allocation methods, and full ECC support for buses between chips inside a POWER5 system for performance and availability.
Figure 2-2 The POWER chip evolution 2.2.4 CMOS, copper, and SOI technology The POWER5 processor design enables IBM Sserver p5 systems to offer clients improved performance, reduced power consumption, and decreased IT footprint size through logical partitioning, Micro-Partitioning and Virtual I/O. It is made using IBM 0.13-...
associated L3 cache and memory are packaged on a single processor book. Access to memory behind another processor is accomplished through the fabric buses. The p5-590 supports up to two processor books (each book is a 16-way) and the p5-595 supports up to four processor books. Each processor book has dual MCMs containing POWER5 processor chips and 36 MB L3 modules.
MCM owns a pair of memory cards. The GX+ interface provides I/O subsystem connection. The output of the lsattr command has been expanded with AIX 5L to include the processor clock rate. IBM Eserver p5 590 and 595 System Handbook...
Table 2-2. The memory locations for each processor chip in the MCMs are illustrated in Figure 2-6 on page 29. Note: DDR1 and DDR2 cannot be mixed within a p5-590/p5-595 server. IBM Eserver p5 590 and 595 System Handbook...
Multichip modules books that contains the POWER5 processors and L3 cache modules Memory cards Service processor unit I/O books that provide the Remote I/O (RIO) ports Fans and blowers for CEC cooling IBM Eserver p5 590 and 595 System Handbook...
Page 53
The main service processor function is located in the CEC. The service processor subsystem runs an independent operating system and directly drives Ethernet ports to connect the external IBM l Hardware Management Console. The service processor unit is not specifically represented in the CEC logic diagram, Figure 2-7 on page 32.
Major design efforts have contributed to the development of the p5-590 and p5-595 to analyze single points of failure within the CEC and to either eliminate them or to provide hardening capabilities to significantly reduce their probability of failure. 2.4.1 CEC backplane The CEC backplane is a double-sided passive backplane that serves as the mounting unit for various system components.
SLOT 2 SLOT 3 SLOT 4 PLUG Seq. Figure 2-10 CEC backplane (front side view) The Central Electronics Complex is an 18 EIA unit drawer that houses: 1 to 4 processors books (nodes). IBM Eserver p5 590 and 595 System Handbook...
Page 57
The processor book contains the POWER5 processors, the L3 cache modules located in Multichip modules, memory and RIO-2 attachment cards. CEC backplane (double sided passive backplane) that serves as the system component mounting unit. Processor books plug into the front side of the backplane.
Smart chips are micro-controllers used to store vital product data (VPD). The smart chip provides a means for securely storing data that cannot be read, altered, or written other than by IBM privileged code. The smart chip provides a means of verifying IBM Sserver On/Off Capacity on Demand and IBM Sserver Capacity Upgrade on Demand activation codes that only the smart chip on the intended system can verify.
capacity and pay for use only when needed. The smart chip is the basis for the CoD function and verifying the data integrity of the data stored in the card. 2.7 I/O drawer The p5-590 and p5-595 use remote I/O drawers (that are 4U) for directly attached PCI or PCI-X adapters and SCSI disk capabilities.
Systems with non-looped configurations could experience degraded performance and serviceability. The system has a non-looped configuration if only one RIO-2 path is running. IBM Eserver p5 590 and 595 System Handbook...
9 (p5-595) 12 (p5-595) 12 (p5-595) On initial orders of p5-590 or p5-595 servers, IBM manufacturing will place dual-loop-connected I/O drawers as the lowest numerically designated drawers followed by any single-looped I/O drawers. A minimum of two cables are required for each loop for each GX+ adapter.
RIO-2 cables, because we use CEC frame, powered 24 inch A frame, and unpowered 24 inch Z frame for the I/O drawer. With the cable ID we calculate the length and the link speed for the RIO-2 cable. IBM Eserver p5 590 and 595 System Handbook...
CEC GX+ I/O RIO-2 PORTS CP A CP B RIO-2 RIO-2 CARD CARD RIO-2 RIO-2 CARD CARD Figure 2-11 Single loop 7040-61D 2.7.4 Half-drawer cabling Although I/O drawers will not be built in half-drawer configurations, they can be cabled to, and addressed by the CEC, in half drawer increments (Figure 2-12). Both STI connectors on one CEC I/O card Px-Cx-T1 and Px-Cx-T2 will be cabled to both ports on P1 RIO-2 Riser card.
Figure 2-12 Dual loop 7040-61D However, to simplify the management of the server we strongly recommend that I/O loops be configured as described in the IBM Sserver Information Center, and to only follow a different order when absolutely necessary. In any case, it becomes extremely important for the management of the system to keep an up-to-date cabling documentation of your systems, because it may be different from the cabling diagrams of the installation guides.
Figure 2-14 blind-swap hot-plug cassette 2.7.6 Logical view of a RIO-2 drawer Figure 2-15 shows a logical schematic of an I/O drawer (FC 5791) and the relationship between the internal controllers, disks, and I/O slots. IBM Eserver p5 590 and 595 System Handbook...
TERM TERM TERM TERM POWER POWER SUPPLY SUPPLY CNTL CNTL CNTL CNTL MIDPLANE SCSI SCSI SCSI SCSI Main Main EADSX EADSX EADSX EADSX EADSX EADSX SS SS SS SS PLANAR BOARD 1 PLANAR BOARD 2 RIO-2 RIO-2 =Soft Switch =SCSI =PCI-X =DIO =PHB...
(BPA) provide both electricity power distribution and reliability, availability, and serviceability infrastructure functions that include: Powering all system components up or down, when requested. These components include I/O drawers and the CEC. IBM Eserver p5 590 and 595 System Handbook...
PHB5 = slots 15, 16, 17, Z1 onboard PHB6 = slots 18, 19, 20, Z2 onboard Figure 2-16 and Figure 2-17 show how to find more information about PCI adapter placement in IBM Sserver Hardware Information Center by placing a PCI placement search for http://publib.boulder.ibm.com/infocenter/eserver/v1r2s/en_US/index.htm...
Search for or refer to: http://publib.boulder.ibm.com/infocenter/eserver/v1r2s/en_US/info/iphak/expansi on61d.htm#expansion61d Figure 2-16 Hardware Information Center search for PCI placement PCI placement Figure 2-17 on page 49 shows sample results after searching for IBM Eserver p5 590 and 595 System Handbook...
Empty slots will be assigned a blind-swap hot-plug cassette with a plastic filler. Additional blind-swap hot-plug cassettes (FC 4599) can be ordered through your IBM Sales Representative or IBM Business Partner. Actual slot numbers are stamped on planar 1 = I1 through I10 (left to right from rear).
2.7.9 Expansion units 5791, 5794, and 7040-61D The following URL for IBM Sserver Information Center provides direction on which adapters can be placed in the 5791, 5794, and 7040-61D expansion units and where adapters should be placed for optimum performance. Figure 2-19 shows the screen capture of the Information Center.
Figure 2-19 PCI placement guide on IBM Sserver Information Center Model 5791 and 5794 expansion units The following is an overview of the PCI placement information located in the Information Center. It is intended to give you an idea of what you may find there.
Page 74
No more than two 1 Gb Ethernet ports per one CPU in a system for maximum performance. More Ethernet adapters may be added for connectivity. Figure 2-20 on page 53 and Figure 2-21 on page 53 show the possible system configurations including the I/O frames. IBM Eserver p5 590 and 595 System Handbook...
A Frame A Frame A Frame Z Frame A Frame A Frame books books books 9-16(18) I/O drawers NOTE: I/O drawers I/O drawers I/O A Frame BPA powers all hardware in I/O Z frame Figure 2-20 Minimum to maximum I/O configuration Figure 2-21 I/O frame configuration example Chapter 2.
Do not use the browser’s back button or the values will not be saved. More information about Advanced System Management Interface (ASMI) can be found in 7.3, “Advanced System Management Interface” on page 175 IBM Eserver p5 590 and 595 System Handbook...
Page 77
Verify the correct value is now displayed in the panel of the unit you are working on. Disconnect all AC power to the unit, wait for the panel to go off, and then reconnect the AC power. Note: The drawer will automatically power on. Log off and close the ASMI and return to the procedure that sent you here.
Page 78
Verify the correct value is now displayed in the panel of the unit you are working on. Log off and close the ASMI. Then return to the procedure that sent you here. IBM Eserver p5 590 and 595 System Handbook...
Virtualization is a critical component in the on demand operating environment, and the system technologies implemented in the POWER5 processor-based IBM Sserver p5 servers provide a significant advancement in the implementation of functions required for operating in this environment. IBM virtualization innovations on the p5-590 and p5-595 provide industry-unique utilization capabilities for a more responsive, flexible, and simplified infrastructure.
Virtualization is a critical component in the on demand operating environment, and the system technologies implemented in the POWER5 processor-based IBM Sserver p5 servers provide a significant advancement in the enablement of functions required for operating in this environment.
a virtual Ethernet adapter and virtual SCSI server. Virtual Ethernet and virtual SCSI are briefly explained in 3.3, “Virtual Ethernet” on page 65 and 3.6, “Virtual SCSI” on page 81. Micro-Partitioning requires the Advanced POWER Virtualization capabilities. 3.2.1 Shared processor partitions The virtualization of processors enables the creation of a partitioning model which is fundamentally different from the POWER4 systems where whole processors are assigned to partitions and are owned by them.
The shared processor partitions are created and managed by the HMC. When you start creating a partition you have to choose between a shared processor partition and a dedicated processor partition. IBM Eserver p5 590 and 595 System Handbook...
Page 83
Virtual processors Virtual processors are the whole number of concurrent operations that the operating system can use. The processing power can be conceptualized as being spread equally across these virtual processors. Selecting the optimal number of virtual processors depends on the workload in the partition. Some partitions benefit from greater concurrence, where other partitions require greater power.
100 percent, and while the machine presents extra capacity not being used, by design the capped partition cannot use it. Time Figure 3-2 Capped shared processor partitions IBM Eserver p5 590 and 595 System Handbook...
Uncapped partitions An uncapped partition has the same definition of a capped partition, except that the maximum limit of processing capacity limit is a soft limit. That means that an uncapped partition may eventually receive more processor cycles than its entitled capacity.
The maximum number of partitions is 254. In a partition, there is a maximum number of 64 virtual processors. IBM Eserver p5 590 and 595 System Handbook...
A mix of dedicated and shared processors within the same partition is not supported. If you dynamically remove a virtual processor you cannot specify a particular virtual CPU to be removed. The operating system will choose the virtual CPU to be removed. Shared processors may render AIX 5L affinity management useless.
Ethernet requires a IBM Sserver p5 590 and 595 system with either AIX 5L Version 5.3 or the appropriate level of Linux and a Hardware Management Console to define the virtual Ethernet devices. Virtual Ethernet does not require the Advanced POWER Virtualization feature because it is a function of the POWER Hypervisor.
Figure 3-4 Example of a VLAN The use of VLAN provides increased LAN security and flexible network deployment over traditional network devices. AIX 5L Version 5.3 virtual LAN support Some of the various technologies for implementing VLANs include: Port-based VLAN Layer 2 VLAN Policy-based VLAN IEEE 802.1Q VLAN...
Page 90
Ethernet adapter. The Virtual I/O Server has a Shared Ethernet Adapter which bridges traffic to the external network. The Shared Ethernet Adapter will be introduced in more detail in 3.4, “Shared Ethernet Adapter” on page 73. IBM Eserver p5 590 and 595 System Handbook...
Figure 3-5 VLAN configuration Interpartition communication Partition 2 and Partition 4 are using the PVID (Port virtual LAN ID) only. This means that: Only packets for the VLAN specified as PVID are received Packets sent are added a VLAN tag for the VLAN specified as PVID by the virtual Ethernet adapter In addition to the PVID the virtual Ethernet adapters in Partition 1 and Partition 3 are also configured for VLAN 10 using specific network interface (en1) create...
Partition 1 / en1 Partition 3 / en1 3.3.2 Virtual Ethernet connections Virtual Ethernet connections supported in POWER5 systems use VLAN technology to ensure that the partitions can only access data directed to them. IBM Eserver p5 590 and 595 System Handbook...
The POWER Hypervisor provides a virtual Ethernet switch function based on the IEEE 802.1Q VLAN standard that allows partition communication within the same server. The connections are based on an implementation internal to the Hypervisor that moves data between partitions. This section will describe the various elements of a virtual Ethernet and implications relevant to different types of workloads.
Ethernet communication is copy bandwidth limited. For more detailed information relating to virtual Ethernet performance considerations refer to the following publication: Advanced POWER Virtualization on IBM Sserver p5 Servers Architecture and Performance Considerations, SG24-5768. 3.3.3 Dynamic partitioning for virtual Ethernet devices Virtual Ethernet resources can be assigned and removed dynamically.
A mixture of virtual Ethernet connections, real network adapters, or both are permitted within a partition. Virtual Ethernet can only connect partitions within a single system. Virtual Ethernet uses the system processors for all communication functions instead of off loading that load to processors on network adapter cards. As a result there is an increase in system processor load generated by the use of virtual Ethernet.
The Shared Ethernet Adapter has the following characteristics: Virtual Ethernet MAC addresses are visible to outside systems Broadcast/multicast is supported ARP and NDP can work across a shared Ethernet IBM Eserver p5 590 and 595 System Handbook...
In order to bridge network traffic between the virtual Ethernet and external networks, the Virtual I/O Server has to be configured with at least one physical Ethernet adapter. One Shared Ethernet Adapter can be shared by multiple VLANs and multiple subnets can connect using a single adapter on the Virtual I/O Server.
An example for this configuration is shown in Figure 3-9. Other aspects which have to be taken into consideration are availability and the possibility to connect to different networks. Figure 3-9 Multiple Shared Ethernet Adapter configuration IBM Eserver p5 590 and 595 System Handbook...
3.4.2 Using Link Aggregation (EtherChannel) to external networks Link aggregation is network port aggregation technology that allows several Ethernet adapters to be aggregated together to form a single pseudo Ethernet device. This technology can be used to overcome the bandwidth limitation of a single network adapter and avoid bottlenecks when sharing one network adapter amongst many client partitions.
Figure 3-10 shows the aggregation of three plus one adapters to a single pseudo Ethernet device including a backup feature. Figure 3-10 Link Aggregation (EtherChannel) pseudo device IBM Eserver p5 590 and 595 System Handbook...
Micro-Partitioning, shared processor pooling, virtual I/O, and Partition Load Manager for AIX 5L logical partitions. The IBM Virtual I/O Server is a special POWER5 partition that provides the ability to implement the virtual I/O function. Virtual I/O enables client partitions to share I/O resources such as Ethernet adapters, SCSI disks, or Fibre Channel disks.
Virtual SCSI and virtual Ethernet are briefly explained in 3.6, “Virtual SCSI” on page 81 and 3.3, “Virtual Ethernet” on page 65. Shared Ethernet functionality is described in 3.4, “Shared Ethernet Adapter” on page 73. IBM Eserver p5 590 and 595 System Handbook...
Although the VIOS is a component of the standard Advanced POWER Virtualization feature, which includes Shared Ethernet Adapter (SEA) service, the VIOS must be installed and configured by the client using the hardware management console (HMC). The VIOS is created using the HMC and is not intended for running applications or for general user logins.
At the time of writing, the Virtual I/O Server supports a maximum of 1024 virtual I/O slots on an IBM Sserver p5 server. A maximum of 256 virtual I/O slots can be assigned to a single partition.
Page 105
CPU cycles. Provided that there is sufficient CPU processing capacity available the performance of virtual SCSI should be comparable to dedicated I/O devices. Chapter 3. POWER5 virtualization capabilities...
Page 106
IBM Eserver p5 590 and 595 System Handbook...
4.1 Capacity on Demand overview IT budgets no longer allow for an infrastructure that can accommodate occasional peaks in demand but whose resources otherwise lie idle. With IBM CoD, you can get the processing or memory resources you need, when you need them, on demand.
Introduced Capacity BackUp The information at the following URL briefly explains the CoD process and features for all IBM Sserver pSeries and IBM TotalStorage products: http://www.ibm.com/servers/eserver/pseries/ondemand/cod/ 4.3 Preparing for Capacity on Demand In this section, a discussion of how to include Capacity on Demand in your enterprise.
There are no restrictions associated with the resale of a systems with inactive CoD resources. However, users of the On/Off CoD offering for Models 590 and 595 are required to terminate this function (at no charge) before the system is resold. IBM Eserver p5 590 and 595 System Handbook...
IBM has established four types of CoD offerings on the p5-590 and p5-595 systems, each with a specific activation plan. Providing different types of CoD offerings gives clients flexibility when determining their resource needs and establishing their IT budgets.
4.5 Capacity BackUp Also available are three new Capacity BackUp features for configuring systems used for disaster recovery. Capacity BackUp for IBM Sserver p5 590 and 595 systems offers an offsite, disaster recovery machine at an affordable price. This disaster recovery machine has primarily inactive Capacity on Demand (CoD) processors that can be activated in the event of a disaster.
Reserve CoD, and Trial CoD resources as well as to enable On/Off CoD. The activation process for IBM Sserver p5 CoD systems is quick and easy. The client simply places an order for a specific CoD feature code and supplies the necessary system configuration data.
Figure 4-2. There is no requirement to set up electronic monitoring of your configuration by IBM except for the On/Off Cod offering that requires a usage report.
Capacity Upgrade on Demand activation procedure IBM will provide the client with a 34-character encrypted key for permanent activation of CUoD resources, both processor and memory, on the p5-595 and p5-590. This key is entered into the system using the hardware management console (HMC) to activate the desired number of additional processors or memory resources.
The system monitors the amount and duration of the activations and generates a usage report which must be sent to IBM monthly. Billing for the activations is then based on the usage report on a quarterly basis. The processor and memory resources are paid for only after they are used, thereby allowing for a flexible, “pay as you go”...
Page 117
1 GB of memory. The system monitors the amount and duration of the activations and generates a usage report which must be sent to IBM monthly. Billing for the activated resources is based on the usage report. Clients only pay for the processor days and memory days after they have been used.
Service Agent or manually. 4. Reporting and order processing and billing The client must report their on/off usage to IBM at least monthly. This usage data will be stored in the TCoD database. This information is used to compute the billing data, which is provided to the sales channel on a quarterly basis.
consecutive hours of testing or production, the result is four processor usage days. If the same client activates two of the previously activated processors in the same 24-hour period, the result is six processor usage days since a new measurement day begins each time processors are activated. Managing On/Off Capacity on Demand resources The hardware management console is used to manage and enable resources temporarily on a p5-590 or p5-595 machine.
The server automatically manages the workload and only charges against the pre-paid Processor Day account when the workload exceeds 100 percent of the base permanently activated processors in the shared processor pool by 10 processor unit. IBM Eserver p5 590 and 595 System Handbook...
Note: Prepaid reserve capacity is supported by the Advanced POWER Virtualization feature and requires an operating system that supports the use of a shared processor pool (AIX 5L Version 5.3). Reserve Capacity on Demand activation procedure Reserve CoD allows clients to implement on/off processor resources using a prepaid, automated plan.
Trial CoD is a complimentary service offered by IBM. Although IBM intends to continue it for the foreseeable future, IBM reserves the right to withdraw Trial CoD at any time, with or without notice.
• Manage On/Off CoD • Manage Reserve CoD • Stop Trial CoD • Shared Pool Utilization – Memory > • Capacity Settings • Manage On/Off CoD • Stop Trial CoD – Show History Log – Show Code Information – Show Billing Information The HMC also has a set of panels that summarize the existing CoD processor and memory capacity settings.
Figure 4-10 CoD Processor Capacity Settings “Trial CoD” HMC window 4.8.1 HMC command line functions In addition to the GUI interface to Capacity on Demand functions on the Hardware Management Console, command line functions are available for most actions. See the HMC documentation regarding use of remote command line functions.
Trial CoD for p5-590 and p5-595 servers is enabled by registering at the pSeries CoD Web site and electronically receiving an activation key. When clients purchase permanent processor activation on p5-590 and p5-595 servers, they IBM Eserver p5 590 and 595 System Handbook...
Software licensing, for the operating system and for IBM middleware as well as for ISV applications is also an important consideration, and IBM has provided...
AIX 5L or selected IBM middleware is executing on. For example, if a client has a 16-way server and is running AIX 5L applications on 12 processors and one of the Linux distributions on the remaining four processors, the client must have 12 license entitlements of AIX 5L.
This value is the one used to determine how many license entitlements are required for the partition. As an example, if an uncapped partition is defined with VP = 7 where the shared processor pool has 12 physical processors assigned, then the number of software license entitlements required for this uncapped partition is seven entitlements.
This data is captured by the system and is reported monthly to IBM. If the report shows 30 processor days were utilized in the quarter, the client must then order 30 daily license entitlements for AIX 5L, and if CSM and HACMP were also running on these temporary processors, the client must order 30 daily license entitlements for CSM and HACMP.
Configuration tools and Chapter 5. rules This chapter is intended to help IBM employees and IBM Business Partners prepare a valid p5-590 and p5-595 configuration to meet a client’s requirement. Section 5.1, “Configuration tools” on page 112 introduces the IBM...
The IBM Configurator tool is an application that provides configuration support for hardware, software, and peripherals associated with IBM product lines that are available for purchase. The IBM Configurator tool allows you to perform the following functions: Configure and upgrade IBM Sserver systems and subsystems...
LVT is a Java™-based tool that is loaded on a Microsoft® Windows® 95 or above workstation with at least 128 MB of free memory. Its footprint on disk, at the time of writing, is about 47 MB. It includes an IBM Java Runtime Environment 1.4. The installation adds an icon to the desktop.
Figure 5-2 LPAR Validation Tool - System Selection dialog Hold your cursor over a field, and additional information is provided, as shown in (Figure 5-3 on page 115). IBM Eserver p5 590 and 595 System Handbook...
Figure 5-3 LPAR Validation Tool - System Selection processor feature selection Partition Specification dialog In the next dialog, partition specifications are entered (Figure 5-4). Figure 5-4 LPAR Validation Tool - Partition Specifications dialog Memory Specification dialog After the Partition Specification fields are complete, the memory specifications are entered for each of the logical partitions you previously specified (Figure 5-5).
This tool answers a common question in planning, “What is the memory usage of the POWER Hypervisor.” LPAR Validation dialog The final dialog enables you to assign features to the various slots defined, as shown in Figure 5-6. IBM Eserver p5 590 and 595 System Handbook...
Figure 5-6 LPAR Validation Tool - slot assignments From this window a detailed report is available, and a validation engine can be selected to point out any errors in the configuration. If changes to the memory configuration are required, you can edit the configuration and change the values in error.
The purpose of this section is to establish the minimum configuration for a p5-590 and p5-595. Appropriate feature codes for each system component are also provided. The IBM Configurator tool will also identify the feature code for each component used to build your system configuration.
Table 5-2 p5-595 minimum system configuration Quantity Component description Feature code (FC) IBM Sserver p5 595 9119-595 16-way, POWER5 Processor Book, FC 7813 or FC 7988 0-way Active Chapter 5. Configuration tools and rules...
Page 142
Adapter for attachment of a device adapter feature code to read CD media or attachment to a NIM server I/O Drawer Attachment Cable FC 6122 Group Slim Line or Acoustic Door Kit FC 6251 or FC 6252 IBM Eserver p5 590 and 595 System Handbook...
LPAR also allows allocation of server resources to various client partitions. IBM Sserver p5 systems introduce an enhanced partitioning model, known as Micro-Partitioning . This feature allows for finer granularity of POWER5 microprocessor allocation across client partitions.
RIO-2 dual loop adapters for attachment of I/O drawers One Multiplexer Card (FC 7812) is required for each processor book to provide a communication channel between the individual processor book and the system service processors IBM Eserver p5 590 and 595 System Handbook...
5.2.4 Memory configuration rules The p5-590 and p5-595 implements a dynamic memory allocation policy that is more sophisticated than the IBM Sserver pSeries POWER4 systems memory allocation model. Unused physical memory can be assigned to a partition without having to specify the precise address of the assigned physical memory in the partition profile;...
Page 146
DDR2 memory is available in 4 GB, 533 MHz fully activated memory cards (FC 7814) Memory resources are configured in the p5-590 and p5-595 servers, using the CEC wizard on the Hardware Management Console. Memory requirements for IBM Eserver p5 590 and 595 System Handbook...
partitions depends upon partition configuration, I/O resources assigned, and applications used. DDR1 memory can be assigned in increments of 16 MB, 32 MB, 64 MB, 128 MB, and 256 MB. The default memory block size varies according to the amount of configurable memory in the system and is summarized in Table 5-3.
Use the following guidelines when attaching I/O drawers to a p5-590 or p5-595 central electronics complex (CEC): Each half of the I/O drawer is powered separately A maximum of eight I/O drawers can be connected to a p5-590 server IBM Eserver p5 590 and 595 System Handbook...
Page 149
A maximum of twelve I/O drawers can be connected to a p5-595 server The p5-590 and p5-595 systems can support the following types of I/O drawers and configurations: A minimum of one I/O drawer (FC 5791 or FC 5794) is required per system –...
4 processor books 5.2.7 Disks, boot devices, and media devices The p5-590 and p5-595 servers must have access to a device capable of reading CD media or to a network installation manager (NIM) server. IBM Eserver p5 590 and 595 System Handbook...
The p5-590 and p5-595 servers incorporate an Early Power Off Warning (EPOW) capability that assists in performing an orderly system shutdown in the event of a sudden power loss. IBM recommends use of the Integrated Battery Backup features or an uninterruptedly power system (UPS) to help ensure against loss of data due to power failures.
Provides additional acoustic dampening for use where a quieter environment is desired. The height of the system frame or Expansion Racks (42U) may require special handling when shipping by air or when moving under a low doorway IBM Eserver p5 590 and 595 System Handbook...
Page 153
Power subsystem The primary system frame always incorporates two bulk power assemblies for redundancy. These provide 350 VDC power for devices located in those frames and associated nonpowered Expansion Racks. These bulk power assemblies are mounted in front and rear positions and occupy the top 8U of the frame. To help provide optimum system availability, these bulk power assemblies should be powered from separate power sources with separate line cords.
Page 154
Expansion Rack for mounting any I/O drawers that cannot be mounted in the primary system frame. clients purchasing p5-595 servers with IBM Eserver p5 590 and 595 System Handbook...
5.2.11 HMC configuration rules In order to configure and administer a partitions on p5-590 and p5-595 systems, you must attach at least one IBM Hardware Management Console for pSeries to the system. We recommend that a second HMC is configured, and each HMC has a second network connection for service and management.
Ethernet ports, two additional PCI-X slots for additional Ethernet adapters, and three USB ports. No charge conversions feature are available for existing IBM 7315-C01/C02/C03/CR2 HMCs from the HMC for POWER4 Licensed Machine Code (FC 0960) to HMC for POWER5 Licensed Machine Code (FC 0961).
Note: It is not possible to connect POWER4 and POWER5 processor-based systems to the same HMC simultaneously. 5.2.12 Cluster 1600 considerations The p5-590 and p5-595 support the IBM eServer Cluster 1600 running Cluster Systems Management V1.4 for AIX 5L Version 5.2, Version 5.3 ( , and 5765-F67) SUSE LINUX Enterprise Server 9 (5765-G16).
Page 158
Given the device bandwidths, it is relatively straightforward to determine the total I/O bandwidth required of the system. However, in addition to the maximum bandwidth that I/O drawers can support, there are adapter limits that must be IBM Eserver p5 590 and 595 System Handbook...
Therefore, you must take a realistic approach. And if you make proper assumptions, you can reach an adequate solution. For more detailed information about capacity and sizing consideration, refer to publication IBM Sserver pSeries Sizing and Capacity Planning: A Practical Guide, SG24-7071. Chapter 5. Configuration tools and rules...
Page 160
IBM Eserver p5 590 and 595 System Handbook...
Diagnose the problem and reconfigure the system as needed. Automatically initiate a repair and call for service. As a result, IBM servers are recognized around the world for their reliable, robust operation in a wide variety of demanding environments. Figure 6-1 on page 141 shows the comprehensive IBM RAS philosophy.
Therefore, the ability to detect imminent failures, and the ability to plan system maintenance in advance are fundamental to reducing outages in a system, and to effectively implement a reliable server. IBM Eserver p5 590 and 595 System Handbook...
The first step to producing a computer that truly has the ability to self heal is to create a highly accurate way to identify and isolate hardware errors. IBM has implemented a server design that builds-in thousands of hardware error check stations that capture and help to identify error conditions within the server.
In this environment, strategically placed error checkers are continuously operating to precisely identify error signatures within defined hardware fault domains. IBM servers are designed so that in the unlikely event that a fatal hardware error occurs, FFDC, coupled with extensive error analysis and reporting firmware in the service processor, should allow IBM to isolate a hardware failure to a single FRU.
6.3.3 Component reliability The components used in the CEC provide superior levels of reliability that are available and undergo additional stress testing and screening above and beyond the industry-standard components that are used in several UNIX OS-based systems today. Fault avoidance is also enhanced by minimizing the total number of components, and this is inherent in POWER5 chip technology, with two processors per chip.
There are several levels of memory protection implemented on the p5-590 and p5-595 systems. From the internal L1 caches to the main memory, several features are implemented to assure data integrity and data recovery in case of memory failures. IBM Eserver p5 590 and 595 System Handbook...
In servers prior to IBMs POWER4 processor-based offerings, this type of error would eventually result in a system crash. The IBM Sserver p5 systems extend the POWER4 technology design and include techniques for handling these errors.
Page 170
These error prone lines cannot then cause system operational problems. In the IBM Sserver p5 product family, the server can dynamically delete up to 10 cache lines. It is not likely that deletion of a couple of cache lines will adversely affect server performance.
The initial state of the array repair data is stored in the FRU vital product data (VPD) by IBM Manufacturing. During the first server IPL, the array repair data from the VPD is used for initialization. If an array fault is detected in an array with redundancy by the array built-in-self-test diagnostic, the faulty array bit is replaced.
POWER Hypervisor can detect the error and direct the service processor to reboot, avoiding other outages. The IBM Sserver p5 595 and 590 include two service processors and two system clocks for added redundancy of components.
Page 173
Figure 6-4 on page 152. Generally, on POWER5 platforms, all PCI adapters controlled by operating system device drivers are connected to a PCI secondary bus created through an IBM designed PCI-PCI bridge. This bridge isolates the PCI adapters and supports hot-plug by allowing program control of the power state of the I/O slot.
RS/6000 and pSeries systems, is the ability for a system to automatically deconfigure an error prone CPU before it causes an unrecoverable system error (unscheduled server outage). It is part of the p5-590 and p5-595 RAS features. IBM Eserver p5 590 and 595 System Handbook...
Page 175
2-way to a 1-way configuration), this limitation was lifted with AIX 5L Version 5.1. AIX 5L Version 5.2 support for dynamic logical partitioning (dynamic LPAR) allowed additional system availability improvements. An IBM Sserver p5 server that includes an inactive CPU (an unused CPU included in a Capacity Upgrade on Demand (CUoD) system configuration) can be configured for CPU hot-sparing.
It is necessary that periodic diagnostics not run against a processor already found to have an error by a current error log entry. CPU Guard provides the required blocking to prevent the multiple logging of the same error. IBM Eserver p5 590 and 595 System Handbook...
6.5.4 Hot-plug components The p5-590 and p5-595 have hot-plug Field Replaceable Units (FRUs). This allows most p5-590 and p5-595 components to be replaced concurrently without the need to power off the system. Parts such as MCMs, L3 cache, memory, and so on, still require a scheduled maintenance window to perform the replacement.
PCI adapters to be removed, added, or replaced without powering down the system. This function enhances system availability and serviceability. IBM introduced blind-swap mechanism to provide concurrent adding or removal of PCI adapters when the system is running. Figure 6-5 shows a basic drawing of the blind-swap hot-plug cassette.
This capability allows the PCI adapters to be concurrently replaced without having to put the I/O drawer into a service position. This technology has been incorporated in selected IBM Sserver p5 servers and I/O drawers.
The resulting commonality makes possible reduced maintenance costs and lower total cost of ownership for IBM Sserver p5 and Sserver i5 systems. This core architecture provides consistent service interfaces and a common approach to service, enabling owners of selected Sserver p5 or Sserver i5 servers to successfully perform set-up, manage and carry out maintenance, and install server upgrades;...
One significant improvement on the IBM Sserver p5 system HMC is to replace the serial attachment method used on predecessor consoles, with a LAN interface allowing high bandwidth connections to servers. With the exception of the p5-590 and p5-595, administrators can choose to establish a private service network, connecting all of their POWER5 processor-based servers and management consoles.
Page 182
The Service call can be directed to the IBM support organization, or to a client identified pager or server identified and set-up to receive service information.
The service processor in the p5-590 and p5-595 product is an improved design when compared to the service processor that was available in the IBM POWER4 processor-based systems. The main service processor function in p5-590 and p5-595 is located in the CEC and is based on the new hardware.
Page 184
The errors are logged in the system nonvolatile random access memory (NVRAM). When the operating system completes booting, the information is passed from the NVRAM into the system error log where it is IBM Eserver p5 590 and 595 System Handbook...
Page 185
Others The service processor also manages the interfaces for connecting Uninterruptible Power Source (UPS) systems to the IBM POWER5 processor-based systems, performing Timed Power-On (TPO) sequences, and interfacing with the power and cooling subsystem. The new service processor incorporates enhanced hardware functions such as an Ethernet service interface, additional serial port communications ports, and larger storage capacity.
To enable service representatives to avoid long lists of repetitive call-home information, the HMC provides the Service Focal Point application. Service Focal serviceable Point recognizes that these errors repeat, and filters them into one event for the service representative to review. IBM Eserver p5 590 and 595 System Handbook...
SFP Graphical User Interface (GUI). From here, the service representative selects a specific fault for repair from a list of open service events; initiating automated maintenance procedures specially designed for the IBM Sserver p5 systems. While it is IBMs plan to provide automated service procedures for each component in the system, at first announce only those components that are concurrently maintainable are supported by the new automated processes.
The difference from the previous versions of AIX and AIX 5L is that the errdaemon used to reset the log file if it was corrupted, instead of repairing it. IBM Eserver p5 590 and 595 System Handbook...
Page 189
System hang detection system hang detection AIX 5L Version 5.1 and higher offers a feature called that provides a mechanism to detect system hangs and initiates a pre-configured action. In some situations, it is difficult to distinguish a system that really hangs (it is not doing any meaningful work anymore) from a system that is so busy that none of the lower priority tasks, such as user shells, have a chance to run.
Hot-plug PCI (future direction) PCI Extended Error recovery (future direction and device driver dependent) To enable Linux to take advantage of the IBM Sserver p5, pSeries and OpenPower enhanced reliability support, a Service Aids Toolkit has been made available for download. This toolkit should greatly enhance Linux availability and serviceability when running on IBM Sserver p5, pSeries and OpenPower systems.
7.1 Service processor functions All IBM Sserver p5 590 and 595 servers are shipped with dual service processor cards. One is considered the primary and the other secondary. They are connected together over the Ethernet ports in the bulk power controller (BPC).
Since there are dual service processors per CEC, both service processors must be updated when firmware updates are applied and activated using the Apply/Activate command on the HMC. The system power code image (stored on the BPC part of service processor) contains the power code for the frames that house the CEC cages, I/O cages, and clustering switches.
ASMI sessions can also be opened while partitions are running. It is possible to connect to the service processor using a personal computer as long as the system is connected to the service processor network. The IBM Sserver Hardware Information Center has a section on accessing the Advanced System Management Interface.
Table 7-1 Table of service processor card location codes Jack ID Location Service Service Function code processor 0 processor 1 System Power Control Network (SPCN) connection J00 CEC Front J01 CEC Front Light Strip connection Light Strip Light Strip J01 CEC Back J00 CEC Back Light Strip Light Strip...
B Side to J01 Bypass BPC-B BYPASS J00D SWTCH BPC-BPC Cross Normal Communication STBY PWR LED UEPO CMPLT LED Figure 7-2 Bulk power controller connections SP and OSC packaging Figure 7-3 Oscillator and service processor IBM Eserver p5 590 and 595 System Handbook...
7.3 Advanced System Management Interface The Web interface to the Advanced System Management Interface is accessible through Microsoft Internet Explorer 6.0, Netscape 7.1, or Opera 7.23 running on a PC or mobile computer connected to the service processor. The Web interface is available during all phases of system operation including the initial program load (IPL) and run time.
Figure 7-4 Select service processor Then, click on the Selected pull-down menu and click on Launch ASM Menu, as illustrated in Figure 7-5. Figure 7-5 Select ASMI IBM Eserver p5 590 and 595 System Handbook...
As part of the Launch ASM Menu dialog, the serial number and hostname/IP address of the service processor are presented and must be verified, as shown in Figure 7-6. Figure 7-6 OK to launch 7.3.2 Accessing ASMI using a Web browser The following instructions apply to systems that are disconnected from a Hardware Management Console (in most cases because of an error), or if you have the need to change the network configuration in the service processor and...
IP addresses are also presented in the welcome pane. This login window will also reveal the licensed internal code (LIC) version in the upper right corner, as shown in Figure 7-7. Figure 7-7 ASMI login IBM Eserver p5 590 and 595 System Handbook...
7.3.4 ASMI user accounts In Table 7-3, you find the default user login and passwords. Usually, you will use the ‘admin login, however, be aware that the client has the ability to change any password. Table 7-3 ASMI user accounts User ID Password admin...
Be aware that the ASMI Reset service processor menu function is comparable to the reset with the pinhole in the operate panel. This is used to recover from an unresponsive service processor. Factory Configuration IBM Eserver p5 590 and 595 System Handbook...
Service processor error log Figure 7-9 provides an example of the service processor error log. You can click on each individual event to obtain a more detailed view of the error. Figure 7-9 ASMI menu: Error /Event Logs Chapter 7. Service processor...
Detailed service processor error log Figure 7-10 provides a detailed view of the error log. Note that it resembles the operating system error log. Figure 7-10 ASMI menu: Detailed Error Log IBM Eserver p5 590 and 595 System Handbook...
Factory Configuration The Factory Configuration menu (Figure 7-11) is used to reset the service processor back to its default factory configuration. Resetting the service processor using this menu option deletes all configuration data such as network configuration information. Figure 7-11 ASMI menu: Factory Configuration 7.3.8 System Configuration ASMI menu The ASMI System Configuration menu contains options to view or modify system-level information including the system name, processing unit identifier,...
This setting specifies whether firmware updates will be done by the HMC or by the operating system. The ASMI menus provide an option to establish this policy, as shown in Figure 7-12. Figure 7-12 ASMI Menu: Firmware Update Policy IBM Eserver p5 590 and 595 System Handbook...
7.3.9 Network Services ASMI menu The Network Services menu contains options to view and modify the network configuration and network access. Figure 7-13 shows the Network Configuration menu. This menu allows a user to change the network configuration. However, changes to the system network configuration must be done when the system is powered off.
Figure 7-14 ASMI menu: Logical Memory Block Size 7.4 Firmware updates IBM will periodically release firmware updates for the p5-590 and p5-595. These updates provide changes to your software, licensed internal code (LIC), or machine code that fix known problems, add new function, and keep your server or Hardware Management Console operating efficiently.
Figure 7-15 Potential system components that require fixes Detailed firmware updating information can be found in the IBM Sserver Hardware Information Center under “Service and support → Customer service and support → Getting fixes” as shown in Figure 7-16. Appendix D, “System documentation”...
Figure 7-16 Getting fixes from the IBM Sserver Hardware Information Center 7.5 System Management Services On the p5-590 and p5-595, the System Management Services (SMS) menus are used to specify the partition boot device. In order to change the boot device options, the user must boot the partition to the System Management Services (SMS) menu.
Figure 7-17 Partition profile power-on properties After the power-on properties in the partition profile have been updated, the boot sequence will pause once the SMS menu is reached. The user will be presented with the SMS main menu, as seen in Figure 7-18 on page 190. The SMS main menu has options for language selection, remote IPL setup, change SCSI settings, select console options, and select boot options.
CA00E1DC (to select console) or CA00E1AE (to select boot option). Note that if the user intends to install the boot image from CD or boot from a diagnostic CD, the user should select the Select Install/Boot Device menu option. IBM Eserver p5 590 and 595 System Handbook...
Figure 7-19 Select Boot Options menu options Figure 7-20 on page 192 shows the Configure Boot Device Order menu options. The boot device order will be stored for the subsequent system boots. Note: The user must check the default boot list before changing the boot list order.
Figure 7-21 on page 193 provides an example of a default boot list. If the default boot list is empty, the user must return to the Configure Boot Device Order menu and select the Restore Default Settings option. IBM Eserver p5 590 and 595 System Handbook...
In a partitioned system, the operator panel shows the word LPAR instead of any partition LED codes. Therefore, all LED codes for system partitions are displayed over the HMC. An imbedded DVD-RAM for creating and storing configuration backup information IBM Eserver p5 590 and 595 System Handbook...
Cluster support when combined with IBM Cluster Systems Management (CSM) V1.4 or later Using a virtual console for each partition or controlled system With this feature, every partition can be accessed over the trusted network HMC connection to the server. This is a convenient feature when the partition is not reachable across the public network.
HMC. The desktop HMC can use a number of IBM displays as shown in e-config. A spare display can be used, however there is no ability to add device drivers to the support nonstandard displays.
When a new system or an upgrade is delivered it will have a document named Start Here for hardware. This document is the starting point for setting up the system and the HMC. A CD named the IBM Sserver Hardware Information Center is included to provide information about the server, the HMC, and other Sserver p5 related topics.
Another network interface would typically be used for the network connection between the HMC and the logical partitions on the managed systems, for the HMC-to-logical partition communications. IBM Eserver p5 590 and 595 System Handbook...
The third interface is an optional additional Ethernet connection that can be used for remote management of the HMC. This third interface can also be used to provide a separate HMC connection to different groups of logical partitions. For example, you could do any of the following: –...
HMC to the service processor are addressed in 7.2, “Service processor cabling” on page 172. Detailed information regarding HMC network cabling can also be found in the IBM Sserver Information Center at the URL listed below; search using the keywords HMC cabling: http://publib.boulder.ibm.com/infocenter/eserver/v1r2s/en_US/index.htm...
HMCs connected to the same hub. One HMC is connected to IBM Service and Support. This HMC is actively managing both the servers. The second HMC is connected but redundant, in that it is not actively managing either of the servers.
This arrangement is probably what most large clients would choose. The HMC is located in the dark machine room. The remote HMC and could be located in the bridge or IT department for operational use. IBM Eserver p5 590 and 595 System Handbook...
Figure 8-3 HMC attached to both private and public network By default, the default DHCP IP address range is 192.168.0.2 - 192.168.255.254. The HMCs network address is 192.168.0.1 and the network mask is 255.255.0.0. Note: With a desktop HMC, by default, only one Ethernet port is provided. To provide the private and open network connections, an Ethernet adapter is required.
9119 system frame BPC-A BPC-B Internal Internal Service Service Processor 1 Processor 0 TCP/IP Ethernet connections Figure 8-4 Primary and secondary HMC to BPC connections IBM Eserver p5 590 and 595 System Handbook...
8.3.4 Predefined HMC user accounts The HMC users (login accounts) are associated with specific definitions of authorized tasks and resources (roles). The default (predefined) roles and users are as follows: super administrator The super administrator acts as the root user, or manager of the HMC system.
Figure 8-5 First screen after login as hscroot user 8.4.1 Required setup information The IBM Sserver Hardware Information Center guides you through cabling up the HMC and then through a checklist to gather information needed to configure the HMC. The information needed includes:...
Both the desktop and rack mounted HMC models come with the pre-installed operating system, HMC code, browser, and IBM Sserver Hardware Information Center pre-loaded. Therefore, the HMC can be unpacked and powered up immediately after the networking components, power, and other basic connections are made.
On the Guided Setup Wizard - Change HMC Data and Time screen (Figure 8-7), enter the correct date, time, and time zone for your environment. This is typically the time zone of the server, assuming the HMC is local to the machine. For IBM Eserver p5 590 and 595 System Handbook...
remote machines you must decide which is the correct time zone for your environment. Figure 8-7 Date and Time settings The Guided Setup Wizard - Change hscroot password screen is now displayed as shown in (Figure 8-8). Enter the current hscroot password (normally this should be the default password of abc123) and then the new password you want.
- where 0 is the number zero rather than the letter o. Enter the new password (minimum 7 characters) you want for the root user ID. The root user is need for maintenance during boot of the HMC such as fsck. IBM Eserver p5 590 and 595 System Handbook...
Figure 8-9 The root password The Create additional HMC users screen will be shown next (no figure is available for this dialog). You can now optionally create new HMC users at this stage. In our example, we decided to create a new hscpe user ID with a role of hmcpe to allow support staff limited access to HMC (for maintenance).
Reboot HMC Press F1 to go to the BIOS setup screen In the IBM Setup Utility, check Network devices are set to enable You can normally leave the LAN adapter speed at automatic detection for the initial setup. This is the best selection for the private service network connection.
But if you use a private indirect connection over a client Ethernet switch or hub, media speed may need to be set (Figure 8-12). Figure 8-12 Speed selection The next panel allows you to specify whether network is private or open (Figure 8-13).
Page 238
Critical data task that you should do regularly on the HMC. If you ever need to recover the HMC (re-install it), the data saved on the backup will be restored including the LAN configuration of the HMC. IBM Eserver p5 590 and 595 System Handbook...
If the managed system is unable to contact the HMC DHCP service, then the managed system will use its last given IP address. If this is the first (or only) HMC connected to the managed servers, select the 192.168.0.2 - 192.168.255.254 range and click Next to continue. If this is the secondary (or redundant) HMC connected to the same p5-590 or p5-595 servers, then you must select a range of addresses which is different from the range which was selected on the first HMC.
LPARs, and the HMC, must use either short-names (hostname only) or long-names (fully qualified domain name) plus the domain name. If you are using DNS, then a fully qualified domain name must be used. IBM Eserver p5 590 and 595 System Handbook...
Figure 8-16 Host name and domain name Next, the default gateway is assigned. In our case we have a private network with direct connection to the service processor and we do not need a default gateway (Figure 8-17). Figure 8-17 Default gateway IP address If we have an open network, you must define a default gateway.
The panel about the configuration of a firewall is next. The client can set up the firewall settings. This is not needed for a basic installation. This ends the IP configuration in the Setup Wizard (Figure 8-19). Figure 8-19 End of network configuration IBM Eserver p5 590 and 595 System Handbook...
Next, you need the client data for the contact information. This information will be included in the problem record (Figure 8-20 and Figure 8-21). If a problem record is called home, the information entered through this dialog is used. Figure 8-20 Client contact information Figure 8-21 Client contact information Chapter 8.
The Guided Setup Wizard - Configure Connectivity to Your Service Provider screen allows you to select the desire communications method used to connect to IBM for service and support related functions. There are four service applications available on the HMC:...
Page 245
IBM. There are three options available to you: Dial-up from the local HMC This option will use the IBM supplied modem with the HMC to dial in to the IBM support network. You may choose this option if your HMC doesn’t have a high speed Internet connection through an open network or has only been configured in a private network.
Next screen is the Agreement for Service Programs and you must review and accept in order to continue (Figure 8-24). Figure 8-24 License agreement Here you must define the modem configuration (Figure 8-25). IBM Eserver p5 590 and 595 System Handbook...
Figure 8-25 Modem configuration Add the phone number for our country or region. You can add more than one number for backup reason (Figure 8-26, Figure 8-27, and Figure 8-28). Figure 8-26 Country or region Chapter 8. Hardware Management Console overview...
After selecting the country, you can add the correct phone number for callhome (Figure 8-27) and other information (Figure 8-28). Figure 8-27 Select phone number for modem Figure 8-28 Dial-up configuration IBM Eserver p5 590 and 595 System Handbook...
Service Applications → Service Agent → Sserver Registration from the HMC desktop (Figure 8-29). If you do not have a valid IBM ID, you can choose to leave this panel blank and manually complete this information later by clicking Service Applications →...
You can choose whether to notify your administrators of only problems reported to IBM (Only call-home problem events) or of all problem events generated. Enter the IP address and port of your SMTP server. Then click the Add button and enter your administrator’s e-mail address and the notification type required.
Number of connected minutes considered a recovery This is the number of minutes after communications is restored between the HMC and the managed system that you want the HMC to wait before considering a recovery successful. Note: The recommended time is two minutes. Number of minutes between outages considered a new incident This is the number of minutes after communication is restored that you want the HMC to wait before considering another outage a new incident.
Figure 8-32 Summary screen Each task which is complete you can see on the status screen (Figure 8-33). It also gives you the option of a status (Figure 2-37). Figure 8-33 Status screen IBM Eserver p5 590 and 595 System Handbook...
If any configuration parts were unsuccessful, you can check the status log for configuration problems. 8.6 HMC security and user management This section discusses security implementation within the HMC environment that includes the following topics: Certificate authority Server security Object manager security HMC User management System Manager Security ensures that the HMC can operate securely in the client-server mode.
This is also available for download from the HMC. To download the Web-based System Management remote client to your Windows OS-based or Linux OS-based PC, type in the following address from your Web Browser: IBM Eserver p5 590 and 595 System Handbook...
#hostname/remote_client.html where hostname is the name of the HMC you are downloading the Web-based System Management remote client from. You choose whether you want the Windows OS-based version or the Linux OS-based version. To download the security package so that the client/server, that is, the PC to HMC, connection is secure, type in the following address in your Web Browser: #hostname/remote_client_security.html (Once again you choose whether you want the Windows OS-based version or...
Inventory Scout provides an automatic configuration mechanism and eliminates the need for you to manually reconfigure Inventory Scout Services. Depending on the levels of your HMC and partition software, you IBM Eserver p5 590 and 595 System Handbook...
Page 257
VPD depending on the resource type. This VPD is sent weekly to a database at IBM by a scheduled job. The initial collection of data will send all collected VPD but any subsequent transmission will only send what has changed since the last transmission to IBM.
Remote support enables connectivity to IBM from the HMC. The Service Agent must to be enabled to allow the HMC to connect to IBM to transmit the inventory of the managed system to IBM. Enabling remote support also allows for the reporting of problems to IBM via the HMC.
Allow automatic problem reporting to your service and support organization Allow remote support center personnel to directly access your server in the event of a problem Ask the client for this information, or fill it out together with the client. IBM Eserver p5 590 and 595 System Handbook...
Figure 8-37 PPP or VPN connection As you can see, this is a VPN connection to IBM 8.8.2 Service Focal Point The Service Focal Point application is used to help service personnel to diagnose and repair problems on partitioned systems. Service personnel use the HMC as the starting point for all hardware service issues.
It also provides a user interface that allows a user to view the events and perform problem determination. Figure 8-38 Open serviceable events IBM Eserver p5 590 and 595 System Handbook...
Figure 8-41 Exchange parts 8.9 HMC service utilities and tasks The Advanced System Management Interface is a service that is provided to allow access to the system to perform actions including powering on the system, changing the boot configuration, and other tasks. ASMI can be launched from a Web browser or from the HMC.
The HMC GUI signon screen will appear. 8.9.2 Determining HMC serial number For some HMC or service processor troubleshooting situations an IBM service representative will have to sign into the HMC. The service password changes daily and is not available for normal client use. If the PE determines a local service engineer can sign on to the HMC the service representative may request the HMC serial number.
FC 2732 - HiPPI FC 6312 - Quad Digital Trunk Telephony PCI FC 4960 - e-business Cryptographic Accelerator FC 4963 - PCI Cryptographic Coprocessor (FIPS-4) Display adapter (maximum GXT135P (16) GXT135P (16) IBM Eserver p5 590 and 595 System Handbook...
Page 269
Statement of Direction for redundant service processor - 1H 2005 More details on interface and adapter dependencies and their effect on these maximums can be found in the IBM Sales Manual or the pSeries PCI Adapter Placement manual available at: http://publib16.boulder.ibm.com/pseries/en_US/infocenter/base/HW_pci_adp...
Rack light indicator LED diagnostics X= Available; - = Not Available; O= Optional Assuming single required I/O drawer; either PCI or PCI-X slot drawer may be installed. Assuming optional I/O drawers are not installed IBM Eserver p5 590 and 595 System Handbook...
Table A-3 Server I/O attachment Server I/O Max. Slots Max. Disk Max. disk Max. I/O Max. disk attachment drawer slots bays bays per drawer capacity drawer system Disk system system drawer capacity system p5-590 9.3 TB FC 5791 20 PCI-X 9.3 TB (internal drawer)
FC 6158 20/40 GB 4 mm Tape Drive FC 6134 60/150 GB 8 mm Tape Drive (internal) FC 6120 80/160 GB VXA Tape Drive (internal) FC 6256 36/72 GB 4mm DAT72 Tape Drive (internal) IBM Eserver p5 590 and 595 System Handbook...
133 MHz 133 MHz 133 MHz Slots 1 through 20 are compatible with PCI or PCI-X adapters Short adapters can go in short or long slots. All slots support Enhanced Error Handling (EEH) IBM Eserver p5 590 and 595 System Handbook...
Note: The Uffff.ccc.sssssss.Pn.Cm..represents the Hardware Management Console (HMC) location code, which provides information as to the identify of the enclosure, backplane, PCI adapter(s), and connector. The ffff.ccc.sssssss in the location code represents the following: ffff = Feature Code of the Enclosure (drawer or node) ccc = the Sequence Number of the Enclosure sssssss = the Serial Number of the Enclosure.
(the processor books and I/O drawers). Front and rear service access is necessary to accommodate the lift tool for servicing of the optional integrated battery backup (IBB). IBM Eserver p5 590 and 595 System Handbook...
Securing the frame The following can be ordered by the client as additional frame-securing options for the 9119-590 and 9119-595. FC 7939 to attach the rack-mounting plates to the concrete floor (non-raised floor) FC 7937 to attach the rack to a concrete floor when on a raised floor (9 1/2 inches to 11 3/4 inches high) FC 7938 to attach the rack to a concrete floor when on a raised floor (11 3/4 inches to 16 inches high)
See Dual power installations for additional information about power. Planning and installation documentation For all information about planning and installation go to the IBM Sserver Hardware Information Center and search for ‘installation planning 595’. The following is provided as an example: http://publib.boulder.ibm.com/infocenter/eserver/v1r2s/en_US/index.htm...
Use one of the following Web sites to find the Information Center that is correct for you. http://publib.boulder.ibm.com/eserver http://publib.boulder.ibm.com/infocenter/eserver/v1r2s/en_US/index.htm On an HMC Click on Information Center and Setup Wizard On a CD-ROM Shipped with the hardware (SK3T-8159) IBM Eserver p5 590 and 595 System Handbook...
As an order from the IBM Publications Center You can order the Information Center through the IBM publication ordering center using the following URL. http://www.ibm.com/shop/publications/order To use the CD you receive with this order, follow the these steps: Install to your PC, mobile computer, LAB drive or a partition on your system If autorun is set on your operating system, the installation program starts when you insert the CD in the CD-ROM drive.
Internet to keep your locally installed version updated. From your PC, follow these steps: Select the Windows Start Menu. Locate in the installed programs list IBM Sserver Hardware Information Center. In this menu, locate Configuration Tools. Select Download Updates from the Internet.
Full-text search is available on the search term that you type in the search field (Figure D-2). Figure D-2 Search field Appendix D. System documentation...
The Navigation bar is in the left frame. List all the categories and topics (Figure D-3). Figure D-3 Navigation bar IBM Eserver p5 590 and 595 System Handbook...
In the toolbar we can customizing the Interface, selecting another language, sending feedback, or getting help. It is located in the upper right corner (Figure D-4). Figure D-4 Toolbar with start off call Appendix D. System documentation...
Ctrl+F (Figure D-5). Figure D-5 Previous pSeries documentation If you have comments for InfoCenter, please send them to us and use the Feedback button on the upper right side. IBM Eserver p5 590 and 595 System Handbook...
Page 296
Institute of Electrical and Interconnect Electronics Engineers PCI Extended Internetwork Protocol (OSI) Processor Host Bridges Initial Program Load PHYP Power Hypervisor Independent Software Vendor Partition Load Manager International Technical POST Power-On Self-test Support IBM Eserver p5 590 and 595 System Handbook...
Page 297
POWER Performance Optimization Secure Socket Layer with Enhanced RISC System Service Point-to-Point Protocol Representative Processor Unit Value SWMA Software Maintenance Agreement Program Temporary Fix SYNC Synchronization PVID Physical Volume Identifier Total Cost of Ownership RAID Redundant Array of Independent Disks TCP/IP Transmission Control Protocol/Internet Protocol...
Page 298
IBM Eserver p5 590 and 595 System Handbook...
IBM Redbooks For information on ordering these publications, see “How to get IBM Redbooks” on page 279. Note that some of the documents referenced here may be available in softcopy only.
Linux Network Installation Service for multiple platforms provided by an AIX 5L Version 5.3 NIM Server Whitepaper http://www-1.ibm.com/servers/eserver/pseries/hardware/whitepapers/network_i nstall.pdf Virtual Networking on AIX 5L Whitepaper http://www-1.ibm.com/servers/aix/whitepapers/aix_vn.pdf IBM Sserver p5 - AIX 5L Support for Micro-Partitioning and Simultaneous Multi-threading Whitepaper http://www-1.ibm.com/servers/aix/whitepapers/aix_support.pdf IBM Sserver Linux on POWER Overview Whitepaper http://www-1.ibm.com/servers/eserver/linux/power/whitepapers/linux_overview .pdf...
CoD activation Web site http://www-912.ibm.com/pod/pod How to get IBM Redbooks You can search for, view, or download Redbooks, Redpapers, Hints and Tips, draft publications and Additional materials, as well as order hardcopy Redbooks or CD-ROMs, at this Web site: ibm.com/redbooks...
Page 302
IBM Eserver p5 590 and 595 System Handbook...
Page 304
170, 172, 175 Capacity on Demand (CoD) 11, 85, 198 admin user 179 activation procedure 91 factory configuration menu 183 Capacity BackUp 90 logical memory block size menu 185 definition of 85 IBM Eserver p5 590 and 595 System Handbook...
Page 305
enablement feature, Capacity Upgrade on De- CoD 36 mand (CUoD) 93 CUoD 23 enablement feature, On/Off Capacity On De- command mand (On/Off CoD) 94 inittab 167 enablement feature, Reserve Capacity on De- lsattr 25 mand (Reserve CoD) 98 lslpp -l 26 enablement feature, Trial Capacity on Demand pmcycles 26 (Trial CoD) 99...
Page 306
FC 2104 252 FC 6122 119–120 FC 2105 252 FC 6134 252 FC 2109 252 FC 6158 252 FC 2732 246 FC 6186 119, 121, 132–133 FC 2738 246, 256 FC 6200 133 IBM Eserver p5 590 and 595 System Handbook...
Page 307
FC 6201 133 FC 8197 124 FC 6203 246 FC 8198 124 FC 6204 246, 257 FC 8398 246 FC 6230 246, 256 FC 8691 131–133 FC 6231 256 Field Replaceable Unit 155 FC 6235 256 field replaceable unit (FRU) 144 FC 6239 246 firewall 209, 220 FC 6240 133...
Page 310
Remote Support Facility 223, 238 simultaneous multi-threading 8, 11, 20, 65 Reserve Capacity on Demand (Reserve CoD) execution unit 20 activation procedure 99 rename registers 20 enablement feature 98 Floating Point registers (FPR) 20 IBM Eserver p5 590 and 595 System Handbook...
Page 311
general-purpose registers (GPR) 20 UPS 146 single point of failure 33, 146 single thread 21 single-bit error correction 147 Virtual Ethernet 58, 80 single-loop 39 ARP 74 single-threaded operation 19, 21 broadcast 74 slimline 260 communication with external networks 70 slot placement 52 interpartition communication 69 SMI-II 26...
Page 312
IBM Eserver p5 590 and 595 System Handbook...
Page 316
The IBM 9119 models 590 BUILDING TECHNICAL Sserver p5 590 and 595 provide an expandable, high-end and 595 INFORMATION BASED ON enterprise solution for managing the computing requirements PRACTICAL EXPERIENCE necessary to become an on demand business.