This information was developed for products and services offered in the U.S.A.
IBM may not offer the products, services, or features discussed in this document in other countries. Consult
your local 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. However, it is the user's responsibility to
evaluate and verify the operation of any non-IBM product, program, or service.
IBM may have patents or pending patent applications covering subject matter described in this document. The
furnishing of this document does not give you any license to these patents. You can send license inquiries, in
writing, to:
IBM Director of Licensing, IBM Corporation, North Castle Drive Armonk, NY 10504-1785 U.S.A.
The following paragraph does not apply to the United Kingdom or any other country where such provisions
are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS
PUBLICATION "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED,
INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT,
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of
express or implied warranties in certain transactions, therefore, this statement may not apply to you.
This information could include technical inaccuracies or typographical errors. Changes are periodically made
to the information herein; these changes will be incorporated in new editions of the publication. IBM may make
improvements and/or changes in the product(s) and/or the program(s) described in this publication at any time
without notice.
Any references in this information to non-IBM Web sites are provided for convenience only and do not in any
manner serve as an endorsement of those Web sites. The materials at those Web sites are not part of the
materials for this IBM product and use of those Web sites is at your own risk.
IBM may use or distribute any of the information you supply in any way it believes appropriate without
incurring any obligation to you.
Any performance data contained herein was determined in a controlled environment. Therefore, the results
obtained in other operating environments may vary significantly. Some measurements may have been made
on development-level systems and there is no guarantee that these measurements will be the same on
generally available systems. Furthermore, some measurement may have been estimated through
extrapolation. Actual results may vary. Users of this document should verify the applicable data for their
specific environment.
Information concerning non-IBM products was obtained from the suppliers of those products, their published
announcements or other publicly available sources. IBM has not tested those products and cannot confirm the
accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the
capabilities of non-IBM products should be addressed to the suppliers of those products.
This information contains examples of data and reports used in daily business operations. To illustrate them
as completely as possible, the examples include the names of individuals, companies, brands, and products.
All of these names are fictitious and any similarity to the names and addresses used by an actual business
enterprise is entirely coincidental.
COPYRIGHT LICENSE:
This information contains sample application programs in source language, which illustrates programming
techniques on various operating platforms. 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 the application programming interface for the operating platform for which the sample
programs are written. These examples have not been thoroughly tested under all conditions. IBM, therefore,
cannot guarantee or imply reliability, serviceability, or function of these programs. 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.
PowerPC®
pSeries®
Redbooks™
Redbooks (logo)™
RS/6000®
Service Director™
TotalStorage®
viiip5-550 Technical Overview and Introduction
Preface
This document is a comprehensive guide covering the IBM Sserver® p5 550 UNIX®
servers. We introduce major hardware offerings and discuss their prominent functions.
Professionals wishing to acquire a better understanding of IBM Sserver p5 products should
consider reading this document. The intended audience includes:
Customers
Sales and marketing professionals
Technical support professionals
IBM Business Partners
Independent software vendors
This document expands the current set of IBM Sserver documentation by providing a
desktop reference that offers a detailed technical description of the p5-550 system.
This publication does not replace the latest IBM Sserver pSeries® marketing materials and
tools. It is intended as an additional source of information that, together with existing sources,
can be used to enhance your knowledge of IBM server solutions.
The team that wrote this Redpaper
This Redpaper was produced by a team of specialists from around the world working at the
International Technical Support Organization, Austin Center.
Giuliano Anselmi is a certified pSeries Presales Technical Support Specialist working in the
Field Technical Sales Support group based in Rome, Italy. For seven years, he was an IBM
Sserver pSeries Systems Product Engineer, supporting Web Server Sales Organization in
EMEA, IBM Sales, IBM Business Partners, Technical Support Organizations, and IBM Dublin
eServer Manufacturing. Giuliano has worked for IBM for 12 years, devoting himself to
RS/6000® and pSeries systems with his in-depth knowledge of the related hardware and
solutions.
Gregor Linzmeier is an IBM Advisory IT Specialist for RS/6000 and pSeries workstation and
entry servers as part of the Systems and Technology Group in Mainz, Germany supporting
IBM sales, Business Partners, and customers with pre-sales consultation and implementation
of client/server environments. He has worked for more than 13 years as an infrastructure
specialist for RT, RS/6000, and AIX® in large CATIA client/server projects.
Wolfgang Seiwald is an IBM Presales Technical Support Specialist working for the System
Sales Organization in Salzburg, Austria. He holds a Diplomingenieur degree in Telematik
from the Technical University of Graz. The main focus of his work for IBM in the past five
years has been in the areas of the IBM Sserver pSeries systems and the IBM AIX operating
system.
Philippe Vandamme is an IT Specialist working in pSeries Field Technical Support in Paris,
France, EMEA West region. With 15 years of experience in semi-conductor fabrication and
manufacturing and associated technologies, he is now in charge of pSeries Pre-Sales
Support. In his daily role, he supports and delivers training to the IBM and Business Partner
Sales force.
The project that produced this document was managed by:
Scott Vetter
IBM U.S.
Thanks to the following people for their contributions to this project:
Ron Arroyo, Lupe Brown, Arzu Gucer, Barb Hewitt, Thoi Nguyen, Jan Palmer, Charlie
Reeves, Craig Shempert, Scott Smylie, Joel Tendler, Ed Toutant, Jane Arbeitman, enley
Jackson, Andy McLaughlin.
IBM U.S.
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
customers.
Your efforts will help increase product acceptance and customer satisfaction. As a bonus,
you'll develop a network of contacts in IBM development labs, and increase your productivity
and marketability.
Find out more about the residency program, browse the residency index, and apply online at:
ibm.com/redbooks/residencies.html
Comments welcome
Your comments are important to us!
We want our papers to be as helpful as possible. Send us your comments about this
Redpaper or other Redbooks in one of the following ways:
Use the online Contact us review redbook form found at:
ibm.com/redbooks
Send your comments in an Internet note to:
redbook@us.ibm.com
Mail your comments to:
IBM Corporation, International Technical Support Organization
Dept. JN9B Building 905 Internal Zip 9053D004
11501 Burnet Road
Austin, Texas 78758-3493
xp5-550 Technical Overview and Introduction
Chapter 1.General description
The IBM Sserver p5 550 deskside and rack-mount servers are designed for greater
application flexibility, with innovative technology, to capitalize on the e-business revolution at
the entry-level or server environments. Introduced with the POWER4 and POWER4+
technology in 2001, available from the 1-way entry-level through the 32-way high-end pSeries
systems, the IBM POWER architecture achieved a new stage of capability characteristics by
including features such as logical partitioning (LPAR). With POWER5 microprocessor
technology, the p5-550 is the first cost-effective, high-performance entry UNIX server to
include the next development of the IBM partitioning concept, Micro-Partitioning.
1
Dynamic LPAR on a 4-way p5-550 allows up to four dedicated partitions. In addition, the
optional Advanced POWER Virtualization feature enables the support up to 40
Micro-Partitions on a 4-way system. Micro-Partitioning technology is an advanced feature of
the POWER5 processor that enables multiple partitions to share a physical processor. The
extended POWER Hypervisor controls dispatching the physical processors to each of the
Micro-partitions. In addition to Micro-Partitioning, the Advanced POWER Virtualization feature
allows sharing of physical network adapters and enables the virtualization of SCSI storage.
In combination with the extraordinary POWER5 processor, Micro-Partitioning is designed to
increase the system management efficiency and lowers the operating expenses by allowing
multiple use of single physical resources installed on the p5-550 system.
Simultaneous multi-threading (SMT) is a standard feature of POWER5 technology that allows
two threads to be executed at the same time on a single processor. SMT is selectable with
dedicated or Micro-partition processors.
The symmetric multiprocessor (SMP) p5-550 system features base 2-way or 4-way, 64-bit,
copper and silicone on insulator (SOI)-based POWER5 microprocessors running at 1.65 GHz
with 36 MB off-chip Level 3 cache configurations. Additional processor configurations are
allowed with the addition of IBM Sserver Capacity on Demand (CoD) features. Main
memory starting at 1 GB can be expanded up to 64 GB, based on the available DIMMs, for
higher performance and exploitation of 64-bit addressing to meet the demands of enterprise
computing, such as large database applications.
Included with the p5-550 are five hot-plug PCI-X slots with Enhanced Error Handling (EEH),
one embedded Ultra320 SCSI dual-channel controller, one 10/100/1000 Mbps integrated
dual-port Ethernet controller, two serial ports, two USB 2.0 capable ports, two HMC ports, two
RIO-2 ports, and two System Power Control Network (SPCN) ports.
The p5-550 includes four front-accessible, hot-swap-capable disk bays in a minimum
configuration with an additional four hot-swap-capable disk bays orderable as an optional
feature. The eight disk bays can accommodate up to 1.17 TB of disk storage using the
146.8 GB Ultra320 SCSI disk drives. Three non hot-swapable media bays are used to
accommodate additional devices. Two media bays only accept slim line media devices, such
as DVD-ROM or DVD-RAM, and one half-height bay is used for a tape drive. The p5-550 also
has I/O expansion capability using the RIO-2 bus, which allows attachment of the 7311 Model
D20 I/O drawers.
Additional reliability and availability features include redundant hot-plug cooling fans and
redundant power supplies. Along with these hot-plug components, the p5-550 is designed to
provide an extensive set of reliability, availability, and serviceability (RAS) features that
include improved fault isolation, recovery from errors without stopping the system, avoidance
of recurring failures, and predictive failure analysis.
2p5-550 Technical Overview and Introduction
1.1 System specifications
Table 1-1 lists the general system specifications of the p5-550 system.
Table 1-1 p5 550 specifications
DescriptionRange
Operating temperature5 to 35 degrees Celsius (41 to 95 F)
Relative humidity8% to 80%
Operating voltage2-way: 100 to 127 or 200 to 240 V AC (auto-ranging)
Operating frequency47/63 Hz
Maximum power consumption1100 watts maximum
MAximum thermal output3754 Btu
a. British Thermal Unit
1.2 Physical package
The following sections discuss the major physical attributes found on a p5-550 in
rack-mounted and deskside versions (Figure 1-1), as shown in Table 1-2. The p5-550 is a 4U,
19-inch rack-mounted system or deskside system depending on the feature code.
4-way: 200 to 240 V AC
a
/hour (maximum)
Table 1-2 Physical packaging of the p5-550
DimensionRack (FC 7886)Deskside (FC 7887)
Height178 mm (7.0 inches)533 mm (21.0 inches)
Width437 mm (17.2 inches)201 mm (7.9 inches)
Depth731 mm (28.8 inches)779 mm (30.7 inches)
Minimum configuration41.4 kg (91 pounds)
Maximum configuration57.0 kg (125 pounds)
Chapter 1. General description 3
Figure 1-1 Rack-mount and deskside versions of the p5-550
1.2.1 IBM eServer p5 550 deskside
The p5-550 is a deskside server that is ideal for environments requiring the user to have local
access to the hardware. A typical example of this would be applications requiring a native
graphics display.
To order a p5-550 system as a deskside version, FC 7887 is required. The system is
designed to be set up by the customer and, in most cases, will not require the use of any
tools. Full set-up instructions are included with the system.
The GXT135P 2D graphics accelerator with analog and digital interfaces (FC 2849) is
available and is supported for SMS, firmware menus, and other low-level functions, as well as
when AIX or Linux starts the X11-based graphical user interface. Graphical AIX system tools
are usable for configuration management if the adapter is connected to the primary console,
such as the IBM L200p Flat-Panel Monitor (FC 3636) or the IBM T541H 15-inch TFT Color
Monitor (FC 3637).
1.2.2 IBM eServer p5 550 rack-mounted
The p5-550 is a 4U rack-mounted server and is intended to be installed in a 19-inch rack,
thereby enabling efficient use of computer room floor space. If the IBM 7014-T42 rack is used
to mount the p5-550, it is possible to place up to 10 systems in an area of
644 mm (25.5 inches) x 1147 mm (45.2 inches).
To order a p5-550 system as a rack-mounted version, FC 7886 is required. In addition to the
rack-mounted version, the p5-550 can be installed in IBM or OEM racks. Therefore you are
required to select one of the following features:
Included with the p5-550 rack-mounted server packaging are all of the components and
instructions necessary to enable installation in a 19-inch rack using suitable tools.
4p5-550 Technical Overview and Introduction
The GXT135P 2D graphics accelerator with analog and digital interfaces (FC 2849) is
available and is supported for SMS, firmware menus, and other low-level functions, as well as
when AIX or Linux starts the X11-based graphical user interface. Graphical AIX system tools
are usable for configuration management if the adapter is connected to a common
maintenance console, such as the 7316-TF3 Rack-Mounted Flat-Panel display.
Figure 1-2 shows the basic ports available on the p5-550.
Figure 1-2 Detailed views of the p5-550 rack-mount system
1.3 Minimum and optional features
The p5-550 system is based on a flexible, modular design, and it features:
Up to two processor books using the POWER5 chip, for a 2-way or 4-way configuration
From 1 to 64 GB of total system memory capacity using DDR1 DIMM technology
Four SCSI disk drives in a minimum configuration, eight SCSI disk drives with an optional
second 4-pack enclosure for an internal storage capacity of 1.17 TB using 146.8 GB
drives
Four PCI-X slots and 1 GX+ slot or five PCI-X slots
Two slim-line media bays for optional storage devices
One half-high bay for an optional tape device
The p5-550, including the service processor (SP) described in 2.11.1, “Service processor
base” on page 36, supports the following native ports:
Two 10/100/1000 Ethernet ports
Two serial ports
Chapter 1. General description 5
Two USB 2.0 ports
Two HMC por ts
Two remote I/O (RIO-2) ports
Two SPCN ports
In addition, the p5-550 features one internal Ultra320 SCSI dual channel controller,
redundant hot-swap power supply (optional), cooling fans, and up to two processor power
regulators.
The system supports 32-bit and 64-bit applications.
1.3.1 Processor card features
The p5-550 accommodates 2-way processor cards with state-of-the-art, 64-bit,
copper-based, POWER5 microprocessors running at 1.65 GHz that share 1.9 MB of L2 on
chip cache, 36 MB of L3 cache, and eight slots for memory DIMMS using DDR1 technology.
This processor card (FC 5237) is a Capacity on Demand (CoD) processor card, which means
it provides the physical processor resources, but an additional activation is needed to use the
processors. Different options for activating them are available; see Table 1-3.
An initial order of the p5-550 system must contain the FC 5237 and at least two processor
activations (2x FC 7871 or FC 7871 and FC8450).
The permanent activation options are Permanent
CUoD (shipped from manufacturing as
activated, or for later activation of available non-activated processors through the HMC) and
no-additional-charge CUoD (one processor at no additional charge, if the initial order
qualifies for the Value Pak options, see Table 1-7 on page 10). Temporary activations of
processors are provided by
“Capacity on Demand” on page 48 for further details about CoD activation concepts.
Table 1-3 contains all the available feature codes for processor cards, at the time of writing.
Table 1-3 Processor card and processor activation feature codes
Each processor card features one POWER5 chip, with two processor cores that share
1.9 MB of L2 cache on-chip, 36 MB of L3 cache, and eight slots for memory DIMMs using
DDR1 memory technology.
Reserve CoD (prepaid 30 day reserve), and On/Off CoD. See 3.1,
8450
7931 (requires one FC 7930 to enable On/Off CoD)
6p5-550 Technical Overview and Introduction
1.3.2 Memory features
The processor cards used in the p5-550 system have eight sockets for memory DIMMs that
must be ordered in quads. Each of the memory features available (at the time of writing) listed
in Table 1-4, therefore, provides four DIMMs.
Table 1-4 Memory feature codes
Feature codeDescription
44441 GB (4x 256 MB) DIMMs, DDR1 SDRAM
44472 GB (4x 512 MB), DIMMs, DDR1
44454 GB (4x 1024 MB) DIMMs, DDR1
44508 GB (4x 2048 MB) DIMMs, DDR1
444916 GB (4x 4096 MB) DIMMs, DDR1
It is recommended that each processor card have an equal amount of memory installed.
Balancing memory across the installed processor cards allows distributed memory accesses
that provide optimal performance. The memory controller will detect a variety of memory
configurations of mixed memory sized DIMMs and DIMMs installed in pairs.
1.3.3 Disk and media features
The minimum p5-550 configuration includes a 4-pack disk drive enclosure. A second 4-pack
disk drive enclosure can be installed by ordering FC 6592 or FC 6593; therefore, the
maximum internal storage capacity is 1.17 TB (using the disk drive features available at the
time of writing). The p5-550 also features two slim-line media device bays and one half-height
media bay. The minimum configuration requires at least one disk drive. Table 1-5 shows the
disk drive feature codes that each bay can contain.
Table 1-5 Disk drive feature code description
Feature codeDescription
327336.4 GB 10 K RPM Ultra3 SCSI disk drive assembly
327736.4 GB 15 K RPM Ultra3 SCSI disk drive assembly
327473.4 GB 10 K RPM Ultra3 SCSI disk drive assembly
327873.4 GB 15 K RPM Ultra3 SCSI disk drive assembly
3275146.8 GB 10 K RPM Ultra3 SCSI disk drive assembly
Any combination of DVD-ROM and DVD-RAM drives of the following devices can be installed
in the two slim-line bays:
DVD-RAM drive, FC 5751
DVD-ROM drive, FC 2640
A logical partition running a supported release of the Linux operating system requires a
DVD-ROM drive or DVD-RAM drive to provide a way to boot hardware diagnostics from CD.
Concurrent diagnostics, as provided by AIX, is not available on Linux at the time of writing.
Supplementary devices can be installed in the half-height media bay, such as:
IBM 80/160 GB Internal Tape Drive with VXA Technology, FC 6120
Chapter 1. General description 7
60/150 GB 16-bit 8 mm Internal Tape Drive, FC 6134
36/72 GB 4 mm Internal Tape Drive, FC 6258
1.3.4 USB diskette drive
For today’s administration tasks, an internal diskette drive is not state-of-the-art. In some
situations, the external USB 1.44 MB diskette drive for p5-550 systems (FC 2591) is helpful.
This super-slim-line and lightweight USB V2 attached diskette drive takes its power
requirements from the USB port. A USB cable is provided. The drive can be attached to the
integrated USB ports, or to a USB adapter (FC 2738). A maximum of one USB diskette drive
is supported per integrated controller/adapter. The same controller can share a USB mouse
and keyboard.
1.3.5 I/O drawers
The p5-550 has five internal PCI-X slots, where four of them are long slots and one is a short
slot. If more PCI-X slots are needed, especially well-suited to extend the number of LPARs
and Micro-Partitions, up to eight Model 7311 Model D20 I/O drawers can be attached to the
rack-mount p5-550 system. Up to four Model D20 drawers can be connected to the two RIO-2
ports on the rear of the system that are provided in a minimum configuration. An additional
four Model D20s can be connected by ordering the Remote I/O expansion card (FC 1806). It
provides two RIO-2 ports located on a interposer card that occupies the short PCI-X slot.
7311 Model D20 I/O drawer
The 7311 Model D20 I/O drawer is a 4U full-size drawer, which must be mounted in a rack. It
features seven hot-pluggable PCI-X slots and optionally up to 12 hot-swappable disks
arranged in two 6-packs. Redundant, concurrently maintainable power and cooling is an
optional feature (FC 6268). The 7311 Model D20 I/O drawer offers a modular growth path for
the p5-550 system with increasing I/O requirements. When a p5-550 is fully configured with
eight attached 7311 Model D20 drawers, the combined system supports up to 61 PCI-X
adapters (in a full configuration, Remote I/O expansion cards are required) and
104 hot-swappable SCSI disks.
PCI-X and PCI cards are inserted from the top of the I/O drawer down into the slot. The
installed adapters are protected by plastic separators, designed to prevent grounding and
damage when adding or removing adapters.
The drawer has the following attributes:
4U rack-mount enclosure assembly
Seven PCI-X slots 3.3 volt, keyed, 133 MHz hot-pluggable
Two 6-pack hot-swappable SCSI devices
Optional redundant hot-plug power
Two RIO-2 ports and two SPCN ports
Note: The 7311 Model D20 I/O drawer initial order, or an existing 7311 Model D20 I/O
drawer that is migrated from another pSeries system, must have the RIO-2 ports available
(FC 6417).
7311 Model D20 I/O drawer physical package
The I/O drawer has the following physical characteristics:
Width: 482 mm (19.0 inches)
8p5-550 Technical Overview and Introduction
Depth: 610 mm (24.0 inches)
Height: 178 mm (7.0 inches)
Weight: 45.9 kg (101 pounds)
Figure 1-3 shows the different views of the 7311-D20 I/O drawer.
Adapters
Service
Access
I/O
Drawer
FrontRear
Operator panel
Power supply 2
Power supply 1
RIO ports
1 2 3 45 6 7
8 9 A B C D 8 9 A B C D
SCSI d isk loc ations and ID s
Reserved ports
SPCN ports
Rack indicator
PCI-X slots
Figure 1-3 7311-D20 I/O drawer
Note: The 7311 Model D20 I/O drawer is designed to be installed by an IBM service
representative.
I/O drawers and usable PCI slots
Only the 7311 Model D20 I/O drawer is supported on a p5-550 system. Depending on the
p5-550 system configuration, the maximum number of I/O drawers supported is different.
Table 1-6 summarizes the maximum number of I/O drawers supported and the total number
of PCI-X slots available.
Table 1-6 Maximum number of I/O drawers supported and total number of PCI slots
p5-550Max number of I/O drawersTotal number of PCI-X slots
Minimum configuration428 + 5
Additional FC 1806856 + 5
1.3.6 Hardware Management Console models
The Hardware Management Console (HMC) provides a set of functions that are necessary to
manage the p5-550 system when LPAR, Capacity on Demand without reboot, inventory and
microcode management, and remote power control functions are needed. These functions
Chapter 1. General description 9
include the handling of the partition profiles that define the processor, memory, and I/O
resources allocated to an individual partition.
The 7310 Model CR2 or the 7310 Model C03 HMCs are specifically for POWER5
processor-based systems. However, an existing 7315 Model CR2 and the 7315 Model C03
(POWER4 processor-based systems HMC) can be converted for POWER5 processor-based
systems use when it is loaded with the HMC software required for POWER5 processor-based
systems (FC 0961).
POWER5 processor-based system HMCs require Ethernet connectivity. Ensure that
sufficient Ethernet adapters are available to enable public and private networks if you need
both. The 7310 Model C03 is a desktop model with only one native 10/100/1000 Ethernet
port, but three additional PCI slots. The 7310 Model CR2 is a 1U, 19-inch rack mountable
drawer that has two native Ethernet ports and two additional PCI slots.
When an HMC is connected to the p5-550, the p5-550 integrated serial ports are disabled. If
you need serial connections, for example, non-Ethernet HACMP heartbeat, you need to
provide an async adapter.
Note: It is not possible to connect POWER4 and POWER5 processor-based systems
simultaneously to the same HMC.
1.4 Value Paks
Value Paks are a new offering available on an initial order only. They provide a predefined
configuration, designed to meet typical customer requirements. Special reduced pricing is
available when a system order satisfies specific configuration requirements for memory, disk
drives, and processors. When a Value Pak is ordered, it is still possible to select additional
features.
If you order a p5-550, you can qualify for a processor activation at no additional charge. The
number of processors, total memory, and quantity and size of disk are the only features that
determine if a customer is entitled to processor activations at no additional charge. You will
also be entitled to a discounted AIX operating system license or you can choose to purchase
the system with no operating system. All Value Paks consist of a standard configuration that
must be part of the order. For each paid processor activation, the customer is entitled to one
processor activationat no additional charge if the following minimum requirements are met:
The system must have at least two disk drives of 73.4 GB or greater each.
There must be at least 2 GB of memory installed for each activated processor, as
described in Table 1-7.
Note: Value Paks are available in an initial order only.
Table 1-7 Value Pak configurations
Value PaksProcessors, FCsMemory (MB)Disk
1.65 GHz2-way, 1x 5237 and 1x 78714096, 2x FC 44472x 73.4 GB (FC 3274)
1.65 GHz4-way, 2x 5237 and 2x 78718192, 2x FC 4445 2x 73.4 GB (FC 3274)
10p5-550 Technical Overview and Introduction
1.5 System racks
The Enterprise Rack Models T00 and T42 are 19-inch wide racks for general use with
IBM Sserver p5, pSeries, and RS/6000 rack-based or rack drawer-based systems. The
racks provide increased capacity, greater flexibility, and improved floor space utilization.
The p5-550 uses a 4U rack-mounted server drawer.
If a p5-550 system is to be installed in a non-IBM rack or cabinet, you should ensure that the
rack conforms to the EIA
1
standard EIA-310-D (see 1.5.6, “OEM rack” on page 14).
Note: It is the customer’s responsibility to ensure that the installation of the drawer in the
preferred rack or cabinet results in a configuration that is stable, serviceable, safe, and
compatible with the drawer requirements for power, cooling, cable management, weight,
and rail security.
1.5.1 IBM RS/6000 7014 Model T00 Enterprise Rack
The 1.8-meter (71 inch) Model T00 is compatible with past and present p5, pSeries, and
RS/6000 racks and is designed for use in all situations that have previously used the older
rack models R00 and S00. The T00 rack has the following features:
36 EIA units (36U) of usable space.
Optional removable side panels.
Optional highly perforated front door.
Optional side-to-side mounting hardware for joining multiple racks.
Standard black or optional white color in OEM format.
Increased power distribution and weight capacity.
An optional reinforced (ruggedized) rack feature (FC 6080) provides added earthquake
protection with modular rear brace, concrete floor bolt-down hardware, and bolt-in steel
front filler panels.
Support for both AC and DC configurations.
DC rack height is increased to 1926 mm (75.8 inches) if a power distribution panel is fixed
to the top of the rack.
Up to four Power Distribution Units (PDUs) can be mounted in the proper bays, but others
can fit inside the rack. See 1.5.3, “AC Power Distribution Unit and rack content” on
page 12.
Optional rack status beacon (FC 4690). This beacon is designed to be placed on top of a
rack and cabled to servers, such as a p5-550, and other components, such as a 7311 I/O
drawer, inside the rack. Servers can be programmed to illuminate the beacon in response
to a detected problem or changes in system status.
A rack status beacon junction box (FC 4693) should be used to connect multiple servers
and I/O drawers to the beacon. This feature provides six input connectors and one output
connector for the rack. To connect the servers or other components to the junction box or
the junction box to the rack, status beacon cables (FC 4691) are necessary. Multiple
junction boxes can be linked together in a series using daisy chain cables (FC 4692).
Weight:
1
Electronic Industries Alliance (EIA). Accredited by American National Standards Institute (ANSI), EIA provides a
forum for industry to develop standards and publications throughout the electronics and high-tech industries.
Chapter 1. General description 11
T00 base empty rack: 244 kg (535 pounds)
T00 full rack: 816 kg (1795 pounds)
1.5.2 IBM RS/6000 7014 Model T42 Enterprise Rack
The 2.0-meter (79.3-inch) Model T42 is the rack that will address the special requirements of
customers who want a tall enclosure to house the maximum amount of equipment in the
smallest possible floor space. The features that differ in the Model T42 rack from the Model
T00 include the following:
42 EIA units (42U) of usable space
AC power support only
Weight:
– T42 base empty rack: 261 kg (575 pounds)
– T42 full rack: 930 kg (2045 pounds)
1.5.3 AC Power Distribution Unit and rack content
For rack models T00 and T42 9-outlet PDUs are available.
PDUs with nine outlets (FC 9176, 9177, 9178, 7176, 7177, and 7178) are available. A T42
rack configured for the maximum number of power outlets would have six PDUs (two
mounted horizontally requiring 2U of rack space), for a total of 54 power outlets.
The p5-550 can be connected to any PDU that is available for the 7014-T00 or 7014-T42
racks.
For detailed power cords requirements and power cord feature codes, see the publication
Site and Hardware Planning Information, SA38-0508. An online copy can be found at:
The first four PDUs ordered for a rack will be mounted vertically in the sides of the rack,
occupying all the four PDU bays available. Any additional PDUs will be mounted horizontally
in the rear of the rack and will occupy 1U of rack space.
Note: Each p5-550, or a system drawer to be mounted in the rack, requires two power
cords, which are not included in the base system order.
The Universal PDU (FC 7188) and the optional Universal PDU to be mounted horizontally in
the rack (FC 9188) will be available on December 31, 2004, supporting a wide range of
country requirements and electrical power specifications. Each Universal PDU provides 12
C13 power outlets for use within a 7014-T00 or 7014-T42 rack, compared to nine C13 power
outlets provided by FC 7176 or FC 7177 PDUs. Nine different power cord features can be
used to connect the PDU to a wall power outlet. Each power cord provides the unique design
characteristics for the different power requirements. To match new power requirements,
these power cords can be requested with an initial order of the rack, or with a later upgrade of
the rack features.
12p5-550 Technical Overview and Introduction
1.5.4 Rack-mounting rules for p5-p550
The primary rules that should be followed when mounting the p-550 into a rack are:
The p5-550 is designed to be placed at any location in the rack. For rack stability, it is
advisable to start filling a rack from the bottom.
Any remaining space in the rack can be used to install other systems or peripherals,
provided that the maximum permissible weight of the rack is not exceeded and the
installation rules for these devices are followed.
Before placing a p5-550 into the service position, it is essential that the rack
manufacturer’s safety instructions have been followed regarding rack stability.
Depending on the current implementation and future enhancements of additional 7311 Model
D20 drawers connected to the p5-550 or single installed p5-550 systems, Table 1-8 shows
examples of the minimum and maximum configurations for different combinations of p5-550s
and attached 7311 Model D20 I/O drawers.
Table 1-8 Minimum and maximum configurations for p550s and 7311-D20s
Only p5-550sOne p5-550,
7014-T00 rack9411
7014-T42 rack10521
one 7311-D20
One p5-550,
four 7311-D20s
One p5-550,
eight 7311-D20s
1.5.5 Additional options for rack
The intention of this section is to highlight some solutions available to provide a single point of
management for environments composed of multiple p5-550 servers or other IBM Sserver
p5, pSeries, and RS/6000 systems.
IBM 7212 Model 102 IBM TotalStorage Storage device enclosure
The IBM 7212 Model 102 is designed to provide efficient and convenient storage expansion
capabilities for select IBM Sserver p5, pSeries, and RS/6000 servers. The IBM 7212 Model
102 is a 1U rack-mountable option to be installed in a standard 19-inch rack using an optional
rack-mount hardware feature kit. The 7212 Model 102 has two bays that can accommodate
any of the following storage drive features:
Digital Data Storage (DDS) Gen 5 DAT72 Tape Drive provides physical storage capacity
of 36 GB (72 GB with 2:1 compression) per data cartridge.
VXA-2 Tape Drive provides a media capacity of up to 80 GB (160 GB with 2:1
compression) physical data storage capacity per cartridge.
DDS-4 tape drive with 20 GB native data capacity per tape cartridge and a native physical
data transfer rate of up to 3 MB/sec, using 2:1 compression so that a single tape cartridge
can store up to 40 GB of data.
DVD-ROM drive is a 5 1/4-inch, half-high device. It can read 640 MB CD-ROM and 4.7 GB
DVD-RAM media. It can be used for Alternate IPL (IBM-distributed CD-ROM media only)
and program distribution.
DVD-RAM drive with up to 2.7 MB/sec throughput. Using 3:1 compression, a single disk
can store up to 28 GB of data. Supported DVD disk native capacities on a single
DVD-RAM disk are as follows: up to 2.6 GB, 4.7 GB, 5.2 GB, and 9.4 GB.
Chapter 1. General description 13
Flat panel display options
The IBM 7316-TF3 Flat Panel Console Kit may be installed in the system rack. This 1U
console uses a 15-inch thin film transistor (TFT) LCD with a viewable area of 304.1 mm x
228.1 mm and a 1024 x 768 picture elements (pels) resolution. The 7316-TF3 Flat Panel
Console Kit has the following attributes:
Flat panel color monitor.
Rack tray for keyboard, monitor, and optional VGA switch with mounting brackets.
IBM Space Saver 2, 14.5-inch keyboard that mounts in the Rack Keyboard Tray and is
available as a feature in 16 language configurations (the track point mouse is integrated
into the keyboard).
Note: It is recommended that you have the 7316-TF3 installed between EIA 20 to 25 of the
rack for ease of use. The 7316-TF3 or any other graphics monitor requires a
POWER GXT135P graphics accelerator (FC 2848 or FC 2849) to be installed in the
server, or other graphic accelerator, if supported.
Hardware Management Console 7310 Model CR2
The 7310 Model CR2 is a 1U, 19-inch rack mountable drawer supported in the 7014 Model
T00 and T42 racks. The 7310 Model CR2 provides one serial port, two integrated Ethernet
ports, and two additional PCI slots. The HMC 7310 Model CR2 has USB ports to connect
USB keyboard and mouse devices.
Note: The HMC serial port can be used for external modem attachment if the Service
Agent call-home function is implemented, and the Ethernet ports are used to communicate
to the service processor in p5-550 systems. An Ethernet cable (FC 7801 or 7802) is
required to attach the HMC to the p5-550 system it controls.
1.5.6 OEM rack
The p5-550 can be installed in a suitable OEM rack, provided that the rack conforms to the
EIA-310-D standard. This standard is published by the Electrical Industries Alliance, and a
summary of this standard is available in the publication Site and Hardware Planning Information, SA38-0508.
The key points mentioned in this standard are as follows:
Any rack used must be capable of supporting 15.9 kg (35 pounds) per EIA unit
(44.5 mm [1.75 inches] of rack height).
To ensure proper rail alignment, the rack must have mounting flanges that are at least
494 mm (19.45 inches) across the width of the rack and 719 mm (28.3 inches) between
the front and rear rack flanges.
It might be necessary to supply additional hardware, such as fasteners, for use in some
manufacturer’s racks.
Figure 1-4 shows the drawing specifications for OEM racks.
14p5-550 Technical Overview and Introduction
Figure 1-4 Reference drawing for OEM rack specifications
Chapter 1. General description 15
16p5-550 Technical Overview and Introduction
Chapter 2.Architecture and technical
overview
This chapter discusses the overall system architecture represented by Figure 2-1. The major
components of this diagram are described in the following sections. The bandwidths provided
throughout this section are theoretical maximums provided for reference. It is always
recommended to obtain real-world performance measurements using production workloads.
The POWER5 chip features single-threaded and multi-threaded execution, providing higher
performance in the single-threaded mode than its POWER4 predecessor at equivalent
frequencies provides. POWER5 maintains both binary and architectural compatibility with
existing POWER4 systems to ensure that binaries continue executing properly and all
application optimizations carry forward to newer systems. The POWER5 provides additional
enhancements such as virtualization, reliability, availability, and serviceability at both chip and
system levels, and it has been designed to support speeds up to 3 GHz.
Figure 2-2 shows the high-level structures of POWER4 and POWER5 processor-based
systems. The POWER4 scales up to a 32-way symmetric multiprocessor. Going beyond 32
processors increases interprocessor communication, resulting in high traffic on the
interconnection fabric bus. This contention negatively affects system scalability.
POWER4POWER5
ProcessorProcessor
Fabric bus
L2
cache
Fabric
controller
L3
cache
Memory
controller
Memory
ProcessorProcessor
Fabric bus
L2
cache
Fabric
controller
L3
cache
Memory
controller
Memory
Fabric bus
L3
cache
ProcessorProcessor
L2
cache
Fabric
controller
Memory
controller
Memory
ProcessorProcessor
L2
cache
Fabric
controller
Memory
controller
Memory
L3
cache
Figure 2-2 POWER4 and POWER5 system structures
Moving the L3 cache provides significantly more cache on the processor side than previously
available, thus reducing traffic on the fabric bus and allowing POWER5 processor-based
systems to scale to higher levels of symmetric multiprocessing. The POWER5 supports a
1.9 MB on-chip L2 cache, implemented as three identical slices with separate controllers for
each. Either processor core can independently access each L2 controller. The L3 cache, with
a capacity of 36 MB, operates as a backdoor with separate buses for reads and writes that
operate at half processor speed. Because of the higher transistor density of the POWER5
130 nm technology, it was possible to move the memory controller on chip and eliminate a
chip previously needed for the memory controller function. These changes in the POWER5
also have the significant side benefits of reducing latency to the L3 cache and main memory
and reducing the number of chips necessary to build a system.
The POWER5 processor supports the 64-bit PowerPC architecture. A single die contains two
identical processor cores, each supporting two logical threads. This architecture makes the
chip appear as a 4-way symmetric multiprocessor to the operating system. The POWER5
processor core has been designed to support both enhanced simultaneous multi-threading
(SMT) and single threaded (ST) operation modes.
The SMT mode maximizes the usage of the execution units. In the POWER5 chip, more
rename registers have been introduced (for floating-point operation, rename registers
increased to 120) that are essential for out of order execution and then vital for the SMT.
18p5-550 Technical Overview and Introduction
Enhanced SMT features
To improve SMT performance for various workload mixes and provide robust quality of
service, POWER5 provides two features:
Dynamic resource balancing
– The objective of dynamic resource balancing is to ensure that the two threads
executing on the same processor flow smoothly through the system.
– Depending on the situation, the POWER5 processor resource balancing logic has
different thread throttling mechanisms.
Adjustable thread priority
– Adjustable thread priority lets software determine when one thread should have a
greater (or lesser) share of execution resources.
– POWER5 supports eight software-controlled priority levels for each thread.
ST operation
Not all applications benefit from SMT. Having threads executing on the same processor will
not increase the performance of applications with execution unit limited performance or
applications that consume all the chip’s memory bandwidth. For this reason, the POWER5
supports the ST execution mode. In this mode, the POWER5 processor gives all the physical
resources to the active thread, allowing it to achieve higher performance than a POWER4
processor-based system at equivalent frequencies. Highly optimized scientific codes are one
example where ST operation is ideal.
Dynamic power management
In current Complimentary Metal Oxide Semiconductor (CMOS) technologies, chip power is
one of the most important design parameters. With the introduction of SMT, more instructions
execute per cycle per processor core, thus increasing the core’s and the chip’s total switching
power. To reduce switching power, POWER5 chips use a fine-grained, dynamic clock gating
mechanism extensively. This mechanism gates off clocks to a local clock buffer if dynamic
power management logic knows the set of latches driven by the buffer will not be used in the
next cycle. This allows substantial power saving with no performance impact. In every cycle,
the dynamic power management logic determines whether a local clock buffer that drives a
set of latches can be clock gated in the next cycle.
In addition to the switching power, leakage power has become a performance limiter. To
reduce leakage power, the POWER5 chip uses transistors with low threshold voltage only in
critical paths. The POWER5 chip also has a low-power mode, enabled when the system
software instructs the hardware to execute both threads at the lowest available priority. In low
power mode, instructions dispatch once every 32 cycles at most, further reducing switching
power. The POWER5 chip uses this mode only when there is no ready task to run on either
thread.
2.1.1 POWER chip evolution
The p5-550 system complies with the RS/6000 platform architecture, which is an evolution of
the PowerPC Common Hardware Reference Platform (CHRP) specifications. Figure 2-3
shows the POWER evolution.
Chapter 2. Architecture and technical overview 19
64bit
32bit
604e
332 /
375
F50
0.18 microns
0.18 microns
SP Nodes
Power3
200+
RS64-II
1.0 to
1.0 to
1.3 GHz
1.3 GHz
Core
Core
262.5
RS64
125
S70
POWER4™
POWER4™
Models 270, B80, and
POWER3 SP Nodes
Power3-II
333 / 375 /
450
F80, H80, M80, S80
RS64-II
340
1.0 to
1.0 to
1.3 GHz
1.3 GHz
Core
Core
H70
0.13 microns
S7A
RS64-III
Copper =
450
pSeries p620, p660,
POWER4+
1.2 to
1.9 GHz
Core
1.2 to
1.9 GHz
Core
RS64-IV
600 / 750
and p680
POWER4
1.0 to 1.3
GHz
p630, p650, p655,
p670, and p690
SOI
+ SOI =
POWER5
POWER5
0.13 microns
0.13 microns
1.5 to
1.5 to
1.9 GHz
1.9 GHz
Core
Core
POWER4+
1.2 to 1.9 GHz
p615, p630,
p650, p655, p670
and p690
Note: Not all
processor speeds
available on all
models
TM
TM
1.5 to
1.5 to
1.9 GHz
1.9 GHz
Core
Core
Shared L2
Shared L2Shared L2
Distributed Switch
Distributed Switch
2001
2001
–Distributed Switch
–Distributed Switch
–Shared L2
–Shared L2
–LPAR
–LPAR
–Autonomic computing
–Autonomic computing
–Chip multiprocessing
–Chip multiprocessing
Figure 2-3 POWER chip evolution
2.1.2 CMOS, copper, and SOI technology
The POWER5 processor design is a result of a close collaboration between IBM Systems and
Technology Group
systems to give customers improved performance, reduced power consumption, and
decreased IT footprint size through logical partitioning. The POWER5 processor chip takes
advantage of IBM leadership technology. It is made using IBM 0.13technology. The POWER5 processor also uses copper and silicon-on-insulator (SOI)
technology to allow a higher operating frequency for improved performance yet with reduced
power consumption and improved reliability compared to processors not using this
technology.
and IBM Microelectronics Technologies that enables IBM Sserver p5
In the p5-550 system, the POWER5 chip has been packaged with the L3 cache chip into a
cost-effective Dual Chip Module (DCM) package. The storage structure for the POWER5 chip
is a distributed memory architecture, which provides high memory bandwidth. Each
processor can address all memory and sees a single shared memory resource. As such, a
single DCM and its associated L3 cache and memory are packaged on a single processor
card. Access to memory behind another processor is accomplished through the fabric buses.
The p5-550 supports up to two processor cards (each card is a 2-way). Each processor card
20p5-550 Technical Overview and Introduction
has a single DCM containing a POWER5 processor chip and a 36 MB L3 module. I/O
connects to the Central Electronic Complex (CEC) subsystem through the GX+ bus. Each
DCM provides a single GX+ bus for a total system capability of two GX+ buses. The GX+ bus
provides an interface to a single device such as the RIO-2 buses, as shown in Figure 2-4.
DI
core
1.9 MB L2
GX+ Controller
Figure 2-4 p5-550 DCM diagram
Each processor card contains a single DCM and the local memory storage subsystem for that
DCM. The processor card also contains LEDs for each FRU
CPU card itself. See Figure 2-5 for a processor card layout view.
DI
core
y
r
o
m
e
M
r
e
l
l
o
r
t
n
o
C
L3 Cache
Data Unit
36 MB
SMI II
SMI II
1
on the CPU card including the
DIMM
DIMM
DIMM
DIMM
DIMM
DIMM
DIMM
DIMM
Figure 2-5 Processor card with DDR1 memory socket layout view
2.2.1 Available processor speeds
The p5-550 operates with a processor clock rate of 1.65 GHz for POWER5 processor-based
systems.
To determine the processor characteristics on a running system, use one of the following
commands:
lsattr -El procX Where X is the number of the processor, for example, proc0 is the first
1
Field Replacement Unit
2
The output of the lsattr command has been expanded with AIX 5L to include the processor clock rate.
processor in the system. The output from the command
Chapter 2. Architecture and technical overview 21
2
would be
similar to the following (False, as used in this output, signifies that the
value cannot be changed through an AIX command interface):
type powerPC_POWER5Processor typeFalse
frequency 165600000Processor SpeedFalse
smt_enabled trueProcessor SMT enabled False
smt_threads 2Processor SMT threads False
state enableProcessor stateFalse
pmcycles -mThis command (in AIX 5L) uses the performance monitor cycle
counter and the processor real-time clock to measure the actual
processor clock speed in MHz. The following is the output of a 2-way
p5-550 running at 1.65 GHz system with SMT enabled:
Note: The pmcycles command is part of the bos.pmapi fileset. First check whether that
component is installed using the lslpp -l bos.pmapi command.
2.3 Memory subsystem
The p5-550 system offers pluggable DIMMs for memory. Each processor card contains eight
slots for up to eight pluggable DIMMs. The minimum memory for a p5-550 processor-based
system is 1 GB and 64 GB as a maximum installable memory option. The p5-550 memory
capacity is dependant on the number of processors ordered for the system and on the type of
memory and functionality required. Figure 2-6 shows the offerings and memory slot
availability.
Cpu 0 runs at 1656 MHz
Cpu 1 runs at 1656 MHz
Cpu 2 runs at 1656 MHz
Cpu 3 runs at 1656 MHz
2.3.1 Memory placement rules
The memory features available at the time of writing for the p5-550 are listed in 1.3.2,
“Memory features” on page 7.
Each memory feature consists of four DIMMs, or quad, and must be installed according to
Figure 2-6. The first quad slots are J0A, J1A, J0C, and J1C, and for the second quad, the
slots are J0B, J1B, J0D, and J1D.
Note: A quad must consist of a single feature (that is made of four identical DIMMs). Mixed
DIMM capacities in a quad will result in reduced RAS.
22p5-550 Technical Overview and Introduction
Figure 2-6 Memory placement for the p5-550
2.3.2 Memory restriction
The p5-550 does not officially support OEM memory, and there is no exception to this rule.
OEM memory is never certified for the use in pSeries and the new p5 servers. If the p5-550 is
populated with OEM memory, you could experience unexpected and unpredictable behavior.
All IBM memory is identified by an IBM logo and a white label printed with a barcode on top
and an alphanumeric string on the bottom, created according to the rule reported in
Figure 2-7.
Figure 2-7 IBM memory certification label
11S1234567YL12345678901FRU P/N EC Level
separator
Seven digits
IBM part number
Twelve digits
IBM plant identificator
and part serial number
Sometimes OEM vendors put a label reporting the IBM memory part number but not the
barcode or the alphanumeric string, or both on their DIMMs.
In case of system failure caused by OEM memory installed in the system, the first thing to do
is to replace the suspected memory with IBM memory and check whether the problem is
corrected. Contact your IBM representative for further assistance if needed.
2.4 System buses
The following sections provide additional information related to the internal buses.
Chapter 2. Architecture and technical overview 23
2.4.1 RIO buses and GX card
Each DCM provides a GX+ bus that is used to connect to an I/O subsystem. In a fully
populated p5-550, there are two GX+ buses, one from each processor card. Each p5-550 has
one GX+ slot with a single GX+ bus. The GX+ slot is not active unless the second processor
card is installed. It is not required for CUoD processor cards to be activated in order for the
associated GX+ bus to be active. The GX+ cards are hot-pluggable. The p5-550 provides two
external RIO-2 ports that can operate up to 1 GHz. An add-in GX adapter card (Remote I/O
expansion card, FC 1806) when plugged into the GX+ slot adds two more RIO-2 ports. The
RIO-2 ports are used for I/O expansion to external I/O drawers. The supported I/O drawer is
7311 Model D20.
2.5 Internal I/O subsystem
The internal I/O subsystem resides on the system planar, and the SP is packaged on a
separate service processor card. Each card is a separate FRU. There is an internal RIO-2
bus imbedded in the system planar. The system planar contains both the Enterprise RIO-2
hub and the PCI-X Host bridge chip to connect to the integrated I/O packaged on the system
planar. Two RIO-2 ports of the Enterprise hub chip are used for the integrated I/O, and the
remaining two ports are routed to external connectors.
The system planar provides five PCI-X slots and several integrated PCI devices that interface
to two EADS-X chips that function as PCI-X to PCI-X bridges to the primary PCI-X buses on
the PCI-X Host bridge chip.
PCI-X slot 5 can accept a short PCI-X or PCI card and its space is shared with the Remote
I/O expansion card. The remaining PCI-X slots are full length cards. The dual 10/100/1000
Mbps Ethernet adapter is integrated on the system planar.
2.5.1 PCI-X slots and adapters
PCI-X, where the X stands for extended, is an enhanced PCI bus, delivering a bandwidth of
up to 1 GB/sec, running a 64-bit bus at 133 MHz. PCI-X is backward compatible, so the
p5-550 systems can support existing 3.3 volt PCI adapters.
The PCI-X slots in the p5-550 system support hot-plug and Extended Error Handling (EEH).
In the unlikely event of a problem, EEH-enabled adapters respond to a special data packet
generated from the affected PCI-X slot hardware by calling system firmware, which will
examine the affected bus, allow the device driver to reset it, and continue without a system
reboot.
64-bit and 32-bit adapters
IBM offers 64-bit adapter options for the p5-550, as well as 32-bit adapters. Higher-speed
adapters use 64-bit slots because they can transfer 64 bits of data for each data transfer
phase. Generally, 32-bit adapters can function in 64-bit PCI-X slots; however, some 64-bit
adapters cannot be used in 32-bit slots. For a full list of the adapters that are supported on the
p5-550 systems, and for important information regarding adapter placement, see the
Sserver.
When a p5-550 is connected to a local area network (LAN), the dual port internal
10/100/1000 Mbps RJ-45 Ethernet controller integrated on the system planar can be used to
accomplish that.
See the Table 2-1 for the list of additional LAN adapters available at the time of writing. IBM
supports an installation with NIM using Ethernet and token-ring adapters (CHRP
platform type).
Table 2-1 Available LAN adapters
Feature codeAdapter descriptionSlotSizeMax
49594/16 Token Ring32 or 64short4
496210/100 Ethernet32 or 64short5
5700Gigabit Ethernet64short5
570110/100/1000 Ethernet64short5
57062-port 10/100/1000 Ethernet64short5
57072-port Gigabit Ethernet - SX64 short5
571810 Gigabit Ethernet PCI-X64short1
2.5.3 Graphic accelerators
The p5-550 supports up to two enhanced POWER GXT135P (FC 2849) 2D graphic
accelerators. The POWER GXT135P is a low-priced 2D graphics accelerator for pSeries and
p5 servers. It can be configured to operate in either 8-bit or 24-bit color modes running at
60 Hz to 85 Hz. This adapter supports both analog and digital monitors. The adapter requires
one short 32-bit or 64-bit PCI-X slot.
3
is the
2.5.4 SCSI adapters
To connect to external SCSI devices, the following adapters provided in Table 2-2 are
available, at the time of writing, to be used in p5-550 system.
Table 2-2 Available SCSI adapters
Feature codeAdapter descriptionSlotSizeMax
5703Ultra320 SCSI RAID, bootable64long4
5712Ultra320 SCSI64short5
6204Ultra SCSI Differential32short2
There is also the option to make the internal Ultra320 SCSI channel externally accessible on
the rear side of the system by installing FC 4273. No additional SCSI adapter is required in
this case. If FC 4273 is installed, a second 4-pack disk enclosure (FC 6592 or FC 6593)
cannot be installed, which limits the maximum number of internal disks to four. FC 4273 also
requires one PCI-X slot.
For more information about the internal SCSI system, see 2.7, “Internal storage” on page 26.
3
CHRP stands for Common Hardware Reference Platform, a specification for PowerPC-based systems that can run
multiple operating systems.
Chapter 2. Architecture and technical overview 25
2.6 Internal serial ports
The serial ports S1 and S2, at the rear of the system, are only available if the system is not
managed using a Hardware Management Console (HMC). In this case, the S1 and S2 ports
support the attachment of serial console and modem.
If an HMC is connected, a
under AIX) and also a modem can be connected to the HMC. The S1 and S2 ports are not
usable in this case.
If additional serial port functionality is needed, optional PCI adapters are available (PCI 8-port
adapter FC 2943 or PCI 128-port adapter FC 2944).
2.7 Internal storage
There is one dual channel Ultra320 SCSI controller managed by the EADS-X chips,
integrated into the system planar, that is used to drive the internal disk drives. The eight
internal drives plug into the disk drive backplane, which has two separate SCSI buses with
four disk drives per bus.
The internal disk drive can be used in two different modes based on whether the SCSI RAID
Enablement Card (FC 5709) is installed (see 2.7.3, “RAID options” on page 27).
The p5-550 supports a split 8-pack disk drive backplane that is designed for hot-pluggable
disk drives. The disk drive backplane docks directly to the system planar. The virtual SCSI
Enclosure Services (VSES) hot-plug control functions are provided by the Ultra320 SCSI
controllers.
2.7.1 Internal media devices
virtual serial console is provided by the HMC (logical device vsa0
The p5-550 provides two slim-line media bays for an optional DVD-ROM (FC 2640) and
optional DVD-RAM (FC 5751), and one media bay for a tape drive. These features are
summarized in Table 2-3.
Table 2-3 Available tape drives
Feature codeDescription
62584-mm 36/72 GB tape (LVD)
61348-mm 60/150 GB tape (LVD)
6120VXA 80/160 GB tape (LVD)
2.7.2 Internal hot swappable SCSI disks
The p5-550 can have up to eight hot-swappable disk drives plugged in the two 4-pack disk
drives backplanes. The hot-swap process is controlled by the SCSI enclosure service (SES),
which is located in the 4-pack disk drives backplane (AIX assigns the name ses0 to the first
4-pack, and ses1 to the second, if present). The two hot-swappable 4-pack disk drives
backplanes can accommodate the devices listed in Table 2-4.
26p5-550 Technical Overview and Introduction
Table 2-4 Hot-swappable disk drive options
Feature codeDescription
327336.4 GB 10,000 RPM Ultra3 SCSI hot-swappable disk drive
327736.4 GB 15,000 RPM Ultra3 SCSI hot-swappable disk drive
327473.4 GB 10,000 RPM Ultra3 SCSI hot-swappable disk drive
327873.4 GB 15,000 RPM Ultra3 SCSI hot-swappable disk drive
3275146.8 GB 10,000 RPM Ultra3 SCSI hot-swappable disk drive
At the time of writing, if a new order is placed with two 4-pack DASD backplanes (FC 6592)
and more than one disk, the system configuration shipped from manufacturing will balance
the total number of SCSI disks between the two 4-pack SCSI backplanes. This is for
manufacturing test purposes, and not because of any limitation. Having the disks balanced
between the two 4-pack DASD backplanes allows the manufacturing process to
systematically test the SCSI paths and devices related to them.
Prior to the hot-swap of a disk in the hot-swappable capable bay, all necessary operating
system actions must be undertaken to ensure that the disk is capable of being deconfigured.
After the disk drive has been deconfigured, the SCSI enclosure device will power-off the slot,
enabling safe removal of the disk. You should ensure that the appropriate planning has been
given to any operating-system-related disk layout, such as the AIX Logical Volume Manager,
when using disk hot-swap capabilities. For more information, see Problem Solving and Troubleshooting in AIX 5L, SG24-5496.
Note: We recommend that you to follow this procedure, after the disk has been
deconfigured, when removing a hot-swappable disk:
1. Release the tray handle on the disk.
2. Pull out the disk assembly a little bit from the original position.
3. Wait up to 20 seconds until the internal disk stops spinning.
4. Now you can safely remove the disk from the 4-pack DASD backplane.
After the SCSI disk hot-swap procedure, you can expect to find SCSI_ERR10 logged in the AIX
error log, with the second word of the sense data equal to 0017. It is generated from a SCSI
bus reset issued by the SES to reset all processes when a drive is inserted, and it is not an
issue.
Hot-swap disks and Linux
Linux does not support the hot-swap of any disk drive at the time of writing; therefore, the
Linux operating system does not support these hot-swappable procedures. A p5-550 system
running Linux must be shut down and powered off before you replace any disk drives.
2.7.3 RAID options
Internal hardware RAID is available on the p5-550. Three options are available:
Install the Dual Channel SCSI RAID Enablement Card (FC 5709). Install four disk drives in
the first 4-pack DASD backplane (FC 6592). This will allow RAID 0, 5, or 10 capabilities
within a single 4-pack of DASD with one RAID controller.
Chapter 2. Architecture and technical overview 27
Install FC 5709. Install a second FC 6592. Install four additional disk drives in the second
4-pack DASD backplane. This will allow RAID 0, 5, or 10 capabilities across two 4-packs
of DASD with one RAID controller.
Install feature number 5709. Install the Ultra320 SCSI 4-Pack Enclosure for Disk Mirroring
(FC 6593). Install the PCI-X Dual Channel Ultra320 SCSI RAID Adapter (FC 5703). Install
the SCSI Cable which connects the PCI Adapter to the second 4-pack DASD backplane
(FC 4267). This will allow RAID 0, 5, or 10 capabilities within each 4-pack of DASD with
two RAID controllers.
Note: Because the p5-550 has eight disk drive slots, customers performing upgrades must
perform appropriate planning to ensure the correct handling of their RAID arrays.
2.8 External I/O subsystem
This section describes the external I/O subsystem, the 7311 Model D20 I/O drawer that is
supported on the p5-550 system.
2.8.1 I/O drawers
As described in Chapter 1, “General description” on page 1 , the p5-550 system has five
internal PCI-X slots, which is enough for average customer configurations. If more PCI-X slots
are needed to dedicate more adapters to a partition or to increase the bandwidth of network
adapters, up to eight 7311 model D20 I/O drawers can be added to the p5-550 system.
The p5-550 system contains a RIO-2 bus to connect the internal PCI-X slots through the
PCI-X to PCI-X bridges and support up to four external I/O drawers. To support more I/O
drawers in one p5-550, a RIO-2 expansion card FC 1806 is needed occupying PCI-X slot 5.
The RIO-2 expansion card supports up to four additional I/O drawers.
2.8.2 7311 Model D20 I/O drawer
The 7311 Model D20 I/O drawer must have the RIO-2 loop adapter (FC 6417) to be
connected to the p5-550 system. The PCI-X host bridge inside the I/O drawer provides two
primary 64-bit PCI-X buses running at 133 MHz. Therefore, a maximum bandwidth of
1 GB/sec is provided by each of the buses. To avoid overloading an I/O drawer, the
recommendation in the
can find it at:
Figure 2-8 on page 29 shows a conceptual diagram of the 7311 Model D20 I/O drawer
subsystem.
IBMSserver Hardware Information Center should be followed. You
28p5-550 Technical Overview and Introduction
PCI-X Host
Bridge
RIO
133 MHz,
64-bit PCI-X
PCI-X Bridge
1
2
6
6
4
4
/
/
1
1
3
3
3
3
133 MHz, 64-bit PCI-X
PCI-X Bridge
4
3
6
6
4
4
/
/
1
1
3
3
3
3
5
6
4
/
1
3
3
6
6
4
/
1
3
3
7
6
4
/
1
3
3
Figure 2-8 Conceptual diagram of the 7311-D20 I/O drawer
7311 Model D20 internal SCSI cabling
A 7311 Model D20 supports hot-swappable disks using two 6-pack disk bays for a total of 12
disks. Additionally, the SCSI cables (FC 4257) are used to connect a SCSI adapter (that can
have various features) in slot 7 to each of the 6-packs, or two SCSI adapters, one in slot 4
and one in slot 7 (see Figure 2-9 on page 29).
Connect the SCSI cable feature to the SCSI adapter
in rightmost slot (7) as shown below:
to 6-pack
backplanes
If a SCSI card is also placed in slot 4, wire as shown below:
to 6-pack
backplanes
Figure 2-9 7311 Model D20 internal SCSI cabling
Note: Any 6-pack and the related SCSI adapter can be assigned to a single partition. If
one SCSI adapter is connected to both 6-packs, both 6-packs can be assigned only to the
same partition.
2.8.3 7311 I/O drawer’s RIO-2 cabling
As described in the 2.8, “External I/O subsystem” on page 28, you can connect up to four I/O
drawers in the same loop and up to eight I/O drawers to the p5-550 system.
SCSI cables FC 4257SCSI cables FC 4257
Chapter 2. Architecture and technical overview 29
Each RIO-2 port can operate at 1 GHz in bidirectional mode and is capable of passing data in
each direction on each cycle of the port. Therefore, the maximum data rate is 4 GB/s per I/O
drawer in double barrel mode.
There is one default primary RIO-2 loop in any p5-550 system. This feature provides two
remote I/O ports for attaching up to four 7311 Model D20 I/O drawers to the system in a single
loop. The optional RIO-2 expansion card can be used to increase the number of I/O drawers
connected to one p5-550 system, and the same rules of the default RIO-2 loop must be
considered. Figure 2-10 shows how you could logically connect four I/O drawers to one
p5-550 system. Actual cables should be routed according to installation instructions.
Cost OptimizedPerformance Optimized
PCI-X slots
I/O draw er #1
I/O draw er #2
I/O draw er #3
I/O draw er #4
Figure 2-10 RIO-2 connections
The RIO-2 cables used have different lengths to satisfy the different connection requirements:
Remote I/O cable, 3.5 m (FC 3147)
Remote I/O cable, 10 m (FC 3148)
2.8.4 7311 Model D20 I/O drawer SPCN cabling
SPCN is used to control and monitor the status of power and cooling within the I/O drawer.
The SPCN is a loop (Figure 2-11 on page 31), the cabling starts from SPCN port 0 on the
p5-550 system to SPCN port 0 on the first I/O drawer. The loop is closed connecting the
SPCN port 1 of the I/O drawer back to the port 1 of p5-550 system. If you have more than one
I/O drawer, you continue the loop connecting the following drawer (or drawers) with the same
rule.
I/O drawer #3
I/O drawer #4
I/O draw er #1
I/O draw er #2
30p5-550 Technical Overview and Introduction
SPCN cabling examples
SPCN cabling examples
Primary drawer
Primary drawer
SPCN port0
SPCN port 0
SPCN port1
SPCN port 1
I/O drawer or secondary drawer
I/O drawer or secondary drawer
SPCN port0
SPCN port 0
SPCN port1
SPCN port 1
Primary drawer
Primary drawer
SPCN port0
SPCN port 0
SPCN port1
SPCN port 1
I/O drawer or secondary drawer
I/O drawer or secondary drawer
SPCN port0
SPCN port 0
SPCN port1
SPCN port 1
I/O drawer or secondary drawer
I/O drawer or secondary drawer
SPCN port0
SPCN port 0
SPCN port1
SPCN port 1
Figure 2-11 SPCN cabling examples
There are different SPCN cables to satisfy different length requirements:
SPCN cable drawer to drawer, 2 m (FC 6001)
SPCN cable drawer to drawer, 3 m (FC 6006)
SPCN cable rack to rack, 6 m (FC 6008)
SPCN cable rack to rack, 15 m (FC 6007)
2.8.5 External disk subsystems
The p5-550 system has internal hot-swappable drives. Internal disks are usually used for the
AIX rootvg and paging space. The two 4-pack backplanes can be split when the SCSI RAID
Enablement Card is not part of the p5-550 configuration, but specific customer requirements
can be satisfied with the several external disks possibilities that the p5-550 supports.
IBM 2104 Expandable Storage Plus
The IBM 2104 Expandable Storage Plus Model DS4 is a low-cost 3U disk subsystem that
supports up to 14 Ultra320 SCSI disks from 18.2 GB up to 146.8 GB, at the time this
publication was written. This subsystem can be used in splitbus mode, meaning the bus with
14 disks could be split into two buses with seven disks each. In this configuration, two
additional LPARs (using dedicated devices) could be provided with up to seven disks for
rootvg by using one Ultra3 SCSI adapter (FC 5712) for each LPAR.
For further information about the IBM 2104 Expandable Storage Plus subsystem, visit the
following Web site:
The IBM TotalStorage FAStT Storage server family consists of five models: Model 100, 600,
700, and 900. The Model 100 is the smallest model, which scales up to 14 TB, and Model 900
is the largest, which scales up to 32 TB of disk storage, at the time this publication was
written. Model 600 provides up to 16 bootable partitions that are attached with the Gigabit
Chapter 2. Architecture and technical overview 31
Fibre Channel adapter (FC 5716). Model 700 provides up to 64 bootable partitions. In most
cases, both the FAStT Storage server and the p5-550 or the 7311 Model D20 I/O drawers are
connected to a storage area network (SAN). If only space for the rootvg is needed, the FAStT
Model 100 is a good solution.
For support of additional features and for further information about the FAStT family, refer to
the following Web site:
The IBM TotalStorage Enterprise Storage Server (ESS) is the high-end premier storage
solution for use in storage area networks. The 2105 Model 800 provides from 582 GB up to
55.9 TB of usable disk capacity. An ESS system can also be used to provide disk space for
booting LPARs or Micro-partitions. An ESS is usually connected to a SAN to which the
p5-550 is also connected by using Gigabit Fibre Channel adapters (FC 5716).
For further information about ESS, refer to the following Web site:
Introduced with the POWER4 processor product line and the AIX 5L Version 5.1 operating
system, the logical partition (LPAR) became available. The technology offered the capability
to divide a pSeries system into separate systems, where each LPAR runs an operating
environment on dedicated attached devices, such as processors, memory, and I/O
components. When the customer requested system flexibility to change the system topology
on demand, it was achieved by modifying the system layout on the required HMC.
Later, dynamic LPAR increased the flexibility, allowing selected system resources, such as
processors, memory, and I/O components to be added and deleted from dedicated partitions
while they are executing. Therefore, AIX 5L Version 5.2 with all the necessary enhancements
to enable dynamic LPAR was introduced in 2002. This requires an attached HMC, with the
proper level of software, to control the system resources and an updated system firmware
level to electronically isolate systems resources. The ability to reconfigure dynamic LPARs
encourages system administrators to dynamically redefine all available system resources to
reach the optimum capacity for each defined dynamic LPAR.
Dynamic logical partitioning is supported by AIX 5L for POWER V5.2, or later. Dynamic
logical partitioning is planned to be supported by SUSE LINUX Enterprise Server 9, or later,
but with reduced functionality (changing memory attributes dynamically is not supported at
the time of writing). Dynamic logical partitioning is not supported by current version of Red
Hat Enterprise Linux AS for POWER Version 3.
USB resources are considered a single group, as are slimline devices. Groups can only be
allocated to a single partition at a time.
2.10 Micro-Partitioning
Based on the partitioning concepts of a stable and well-known mainframe technology and
existing LPAR/dynamic LPAR implementation on POWER4 and POWER4+ servers, The
32p5-550 Technical Overview and Introduction
POWER5 systems introduce an enhanced partitioning model available as a hardware
feature.
The Micro-Partitioning model offers a virtualization method of system resources. In POWER5
processor-based systems, physical resources are abstracted into virtual resources that are
available to partitions. This sharing method is the primary feature of this new partitioning
concept and it happens automatically after a initial system setup.
POWER5 Micro-Partitioning specifies processor capacity in processing units. One processing
unit represents 1% of one physical processor. A Micro-partition defined with 220 processing
units is equivalent to the power of 2.2 physical processors. Creating a micro-partition, the
minimum capacity is 10 processing units, or 1/10 of a physical processor. A maximum of 10
micro-partitions for each physical processor can be defined, but on a loaded system the
practicle limit is less.
Micro-partitions can also be defined with the capped and uncapped attributes. A capped
micro-partition is not allowed to exceed the defined share, while an uncapped partition is
allowed to consume additional capacity with fewer restrictions. Uncapped partitions can be
configured to the total idle capacity of the server or a percentage of it. Configuration through
the HMC menus sets the allowed share and the capped or uncapped attribute.
The POWER5 processor-based systems use the POWER Hypervisor, which is the new
active Hypervisor to execute the Micro-partition model. The Hypervisor of existing POWER4
processor-based systems is working on a demand basis, as the result of machine interrupts
and callbacks to the operating system.
The Advanced POWER Virtualization Feature (Virtual I/O Server) described in 2.10.1,
“Advanced POWER Virtualization feature” on page 34, facilitates the understanding of all the
POWER5 and POWER Hypervisor enhancements to reach the highest level of granularity of
installed system resources.
Figure 2-12 shows the Micro-partition LPAR concept.
Micro-Partitioning
network
8 CPUs2 CPUs 4 CPUs 4 CPUs6 CPUs
Virtual I/O
Server
External
storage
AIX v5.3
Virtual
SCSI
Virtual
Adapter
I/O
Storage Network
LinuxAIX
v5.2
Virtual Ethernet
POWER Hypervisor
I/O
Sto Net
I/O
Sto Net
Sto Net
AIX
v5.3
I/O
AIX v5.3AIX v5.3
LinuxLinux
HMC
LinuxLinux
AIX v5.3AIX v5.3
AIX v5.3AIX v5.3
I/O
S N
Figure 2-12 Micro-partition LPAR
Chapter 2. Architecture and technical overview 33
2.10.1 Advanced POWER Virtualization feature
The Advanced POWER Virtualization feature is a hardware feature (FC 7941 for the p5-550)
that is available as an option on all IBM Sserver POWER5 processor-based systems. The
Advanced POWER Virtualization feature is a combination of hardware enablement for
Micro-partitions and software that supports the Virtual I/O environment on POWER5
processor-based systems.
The Advanced POWER Virtualization feature includes:
Firmware enablement for Micro-partitions
Installation image for the Virtual I/O server software that supports:
– Ethernet adapter sharing
– Virtual SCSI server
Partition Load Manager
2.10.2 Virtual I/O server
The Virtual I/O server is a special purpose partition to provide virtual I/O resources to client
partitions. The Virtual I/O server owns the real resources that will be shared with the other
clients. The Virtual I/O technology allows a physical adapter assigned to a partition to be
shared by one or more partitions, enabling clients to minimize their number of physical
adapters. Virtual I/O will be used to reduce costs by eliminating the requirement that each
Micro-partition has a dedicated network adapter, disk adapter, and disk drive.
It is preferred that you use the Virtual I/O server in a partition with dedicated resources to help
ensure stable performance.
Three major functions are provided from the Virtual I/O Server, as described in the following
sections.
Virtual Ethernet
To enhance intercommunication between partitions, either dedicated partitions or
Micro-partitions, Virtual Ethernet allows in-memory connections at a high bandwidth from
partition to partition. Virtual Ethernet working on LAN technology allows a transmission speed
in the range of 1 to 3 GB/sec depending on the MTU
to 256) are supported in a partition, where a single Virtual Ethernet resource can be
connected to another Virtual Ethernet, a real network adapter, or both in a partition.
4
size. Virtual Ethernet connections (up
Shared Ethernet adapter
The Shared Ethernet adapter is a new service that acts as a layer 2 network switch to route
network traffic from a Virtual Ethernet to a real network adapter. The Shared Ethernet Adapter
must run in a Virtual I/O Server partition.
The advantages of the virtual Ethernet services is that partitions to communicate outside the
system without having a physical network adapter attached to the partition. At the time of
writing, up to 16 Virtual Ethernet x 18 VLANs can be shared on a single network interface.
The amount of network traffic will limit the number of client partitions served through a single
network interface.
4
Maximum Transmission Unit
34p5-550 Technical Overview and Introduction
Virtual SCSI
Access to real storage devices is implemented through the Virtual SCSI services, a part of the
Virtual I/O Server partition. Logical volumes created and exported on the Virtual I/O Server
partition will be shown at the Virtual Storage Client partition as a SCSI disk. All current
storage device types, such as SAN, SCSI, and RAID, are supported. iSCSI and SSA are not
supported.
Note: The Shared Ethernet adapter and Virtual SCSI server function is provided in the
Virtual I/O Server that is included in the optional Advanced POWER Virtualization feature.
2.10.3 Partition Load Manager
The Partition Load Manager (PLM) provides automated processor and memory distribution
between dynamic LPAR and Micro-partition-capable logical partition running AIX 5L. The
PLM application is based on a client/server model to share system information, such as
processor or memory events, across the concurrent present logical partitions.
To improve the overall resource utilization of a partitioned system, PLM uses user-defined
resource management policies to determine the additional resources, such as processors
and memory, for each requesting partition.
PLM uses the Resource Monitoring and Control (RMC) subsystem for network
communication, which provides several events on every managed Micro-partition node. The
following events are registered on all managed Micro-partition nodes:
Memory-pages-steal high thresholds and low thresholds
Memory-usage-high thresholds and low thresholds
Processor-load-average high threshold and low threshold
To ensure a secure communication between managed Micro-partition nodes, OpenSSH and
Kerberos V5 are supported in PLM to have a secure communication and an authentication
mechanism for administrators. If Kerberos is not installed, PLM uses the next configured
authentication method, such as AIX authentication.
2.10.4 Simultaneous multi-threading
As a permanent requirement for performance improvements at the application level,
simultaneous multi-threading (SMT) functionality is embedded in the POWER5 chip
technology. Developers are familiar with process-level parallelism (multi-tasking) and
thread-level parallelism (multi-threads). SMT is the next stage of processor saturation for
throughput-oriented applications to introduce the method of instruction-level parallelism to
support multiple pipelines to the processor.
By default, SMT is activated. On a 2-way POWER5 processor-based system, the operating
system views the available processors as a 4-way system. To achieve a higher performance
level, SMT is also applicable in Micro-Partitioning, capped or uncapped, and dedicated
partition environments.
Simultaneous multi-threading is supported on POWER5 systems running AIX 5L V5.3 or
Linux operating system-based at a required 2.6 kernel. AIX provides the smtctl command
that turns SMT on and off without a subsequent reboot. For Linux, an additional boot option
must be set to activate SMT after a reboot.
Chapter 2. Architecture and technical overview 35
2.11 Service processor
The service processor (SP) is an embedded controller based on a PowerPC 405GP
processor (PPC405) implementation running the SP internal operating system. The SP
operating system contains specific programs and device drivers for the SP hardware
(Figure 2-13 on page 36).
The p5-550 uses the SP implementation. The key components include a FSP-Base (FSP-B)
and an Extender chipset (FSP-E). FSP-B and FSP-E are implemented on a dedicated card.
Rear serial
ports system
S1
S2
Rear ports
FSP
HMC1
HMC2
SPCN1
SPCN2
Figure 2-13 Service processor logical diagram
2.11.1 Service processor base
System planar
FSP-E
UART
#3
UART
#4
MUX
UART#2UART
RJ45
Ethn.1
Ethn.2
FSP-B
RJ45
#1
Flash
NVRAM
DDR
Interface
Flash ROM
48 MB
ALE
NVRAM
controller
SRAM
1 MB
DDR
64 MB
The PPC405 core is 5-stage pipeline instruction processor and contains 32-bit general
purpose registers. The Flash ROM contains a compressed image of a software load.
The SP base unit offers the following connections:
Two Ethernet Media Access Controller3 (MAC3) cores, which is a generic implementation
of the Ethernet Media Access (MAC) protocol compliant with ANSI/IEEE 802.3, IEEE
802.3u, ISO/IEC 8802.3 CSMA/CD standard. The Ethernet MAC3 supports both half
duplex (CSMA/CD) and full duplex operation at 10/100 Mbps. Both Ethernet port are
visible only to the service processor.
Two serial interfaces, accessible only though the serial ports of p5-550 on the rear side. At
the time of writing, the System Management Interface (SMI) is usable if a connection is
established to serial port 1. Terminals connected to serial port 2 receive only boot
sequence information without manual interaction. When the HMC is connected to the SP,
the serial ports are disabled and do not provide any external connection.
36p5-550 Technical Overview and Introduction
2.11.2 Service processor extender
The SP extender unit offers two system power control network (SPCN) ports that are used to
control the power of the attached I/O subsystems. The SPCN control software and the service
processor software are run on the same PPC405 processor.
2.12 Boot process
From the earlier RS/6000 systems, through the previous pSeries systems, the boot process
passed through several enhancements. With the implementation of the POWER5 chip
technology in the pSeries platform, the boot process is enhanced for the flexibility that the
POWER5 processor-based hardware features. Depending on the customer configuration, a
system may or may not require the use of an HMC to manage the system. The boot process,
based on the Initial Program Load (IPL) setup, will depend on the hardware setup and on the
way we will use the features that POWER5 processor-based systems provide.
The IPL process starts when power is connected to the system. Immediately after, the SP
starts an internal self test based on integrated diagnostic programs (Built-In-Self-Test, BIST).
Only if all the test units have been successfully passed, the system status changes to
standby.
2.12.1 IPL flow without an HMC attached to the system
When the system status is standby, the SP provides a System Management Interface (SMI)
that can be accessed by pressing any key on an attached serial console keyboard, or the
Advanced System Management Interface (ASMI) using a Web browser
that is connected to the SP on an Ethernet network.
5
on a client system
The SP and the ASMI are standard on all POWER5 processor-based hardware. Both system
management interfaces require you to enter the general or admin ID password and allow you
to set flags that affect the operation of the system, according to the provided password, such
as auto power restart, to view information about the system (such as the error log and VPD),
network environment access setup, and to control the system power.
You can start and shut down the system in addition to setting IPL options. This server has a
permanent firmware boot side, or A side, and a temporary firmware boot side, or B side. New
levels of firmware should be installed on the temporary side first in order to test the update’s
compatibility with your applications. When the new level of firmware has been approved, it
can be copied to the permanent side.
In the SMI and ASMI, you can view and change the following IPL settings:
System boot speed.
Fast or Slow. Fast boot results in skipped diagnostic tests and shorter memory tests
during the boot.
Firmware boot side for next boot.
Permanent or Temporary. Firmware updates should be tested by booting from the
temporary side before being copied into the permanent side.
System operating mode.
5
Supported browsers are Netscape (Version 7.1), Microsoft Internet Explorer (version 6.0), and Opera (Version
7.23). At the time of writing, previous versions of these browsers are not supported. JavaScript™ and cookies must
be enabled.
Chapter 2. Architecture and technical overview 37
Manual or Normal. Manual mode overrides various automatic power-on functions, such as
auto-power restart, and enables the power switch button.
AIX/Linux partition mode boot, available only if the system is not managed by the HMC.
– Service mode boot from saved list. This is the preferred way to run concurrent AIX
diagnostics.
– Service mode boot from default list. This is the preferred way to run stand-alone AIX
diagnostics.
– Boot to open firmware prompt.
– Boot to System Management Service (SMS) to further select the boot devices or
network boot options.
Boot to server firmware:
Select the state for the server firmware: Standby or Running.
When the server is in the server firmware standby state, partitions can be set up and
activated
2.12.2 Hardware Management Console
Depending on the model, the HMC provides a number of native serial ports and Ethernet port.
One serial port can be used to attach a modem for the Service Agent. The Service Agent
Connection Manager can be used instead if the HMC has a TCP/IP port 80 connection to the
Internet. The HMC provides an Ethernet port (or ports) to connect to partitions on its
POWER5 processor-based managed systems. The network connection is mandatory for the
HMC to p5 systems, and highly recommended between the HMC and partitions. It supports
the following functions:
Logical partition configuration and management
Dynamic logical partitioning
Capacity and resource management
System status
HMC management
Service functions (for example, Microcode Updates and Service Focal Point)
Remote HMC interface
Note: The same HMC cannot be attached to POWER4 and POWER5 processor-based
systems simultaneously, but for redundancy purposes, one POWER5 processor-based
server can be attached to two HMCs.
All the managed servers must be authenticated from the HMC. If a new attached system is
discovered, the HMC will prompt you to set two passwords using the HMC interface:
Advanced System Management general user ID password
Advanced System Management admin ID password
HMC access password
2.12.3 IPL flow with an HMC attached to the system
When the system status is in standby, you can use the HMC to open a virtual terminal and
access the SMI, or launch a Web browser to access the ASMI.
38p5-550 Technical Overview and Introduction
Using the SMI or the ASMI, you can view or modify the proper IPL settings in order to set the
boot mode to partition standby and then turn the system on. However, the HMC can be also
used to power on the managed system (and is highly recommended). Using the HMC to turn
the system on requires you to select one of the following choices:
Partition Standby
– The Partition Standby power-on mode allows you to create and activate logical
partitions.
– When the Partition Standby power-on is completed, the operator panel on the
managed system displays LPAR..., indicating the managed system is ready for you to
use the HMC to partition its resources and, possibly, activate them.
– When a partition is activated, the HMC requires you to select the boot mode of the
single partition.
System Profile
The System Profile option powers on the system according to a predefined set of profiles.
The profiles are activated in the order in which they are shown in the system profile.
Partition autostart
This option powers on the managed system to partition standby mode and then activates
all partitions that have been designated autostart.
After the system boots with any of the above choices, the HMC can be used to manage the
system, such as continuing to boot from the operating system or manage the logical
partitions. See 2.12.2, “Hardware Management Console” on page 38.
2.12.4 Definitions of partitions
Describing the detailed process to work with the HMC and the management tasks to create
and manage a logical partition (LPAR or dynamic LPAR) is not the intention of this
documentation. The following section describes the additional functionality used to create
partitions that are using fractional elements of available system resources, namely
Micro-Partitioning.
For a better understanding of the partitioning concept, this section contains an overview of
common terminology. There are two important components:
Managed systems,
profiles
Managed systems
Managed systems are physical systems that are managed by the HMC, whereby one HMC
can manage more managed systems at a time.
Profiles
A profile defines the configuration of a logical partition or managed system. There are three
types of profiles that can be used to create multiple profiles for each logical partition or
managed system:
Partition profile
– A partition profile includes the collection of resource specifications, such as processing
units, memory, and I/O resources, because a logical partition does not know any
resource until it is activated.
Chapter 2. Architecture and technical overview 39
– A logical partition can have more than one partition profile, bu at least one as a
minimum requirement.
All resources dedicated partition profile
A partition profile that contains the entire resource list of the machine, using all physical
resources working as one system.
System profile
– A system profile is an ordered list of partition profiles. When you activate a system
profile, the managed system will attempt to activate the partition profiles in the defined
order.
– To enhance the flexibility to use the system within several different logical
configuration, a system profile can be defined to collect more than one partition profile
to provide the requested system behavior.
2.12.5 Hardware requirements for Micro-Partitioning
To implement the Micro-partition concept on a POWER5 processor-based system, resource
planning is important to have a base configuration and enough flexibility to make desirable
changes to the running logical partitions. To configure a Micro-partition, the minimum
requirements needed are processors, memory, and possibly an expansion unit to define more
Micro-partitions than possible in a single system.
Processors
Within POWER5 technology, and depending on performance requirements, a logical partition
can be created by using a shared processor pool or a dedicated processor.
Shared processors can be defined by a fractional number of 1/10 as minimum requirement of
a real processor. To calculate the required processor power, a real processor is divided in
100 processing units, and 1/10 of a processor is equal to 10 processing units.
Dedicated processors are entire processors that can be assigned to a single logical partition
without the capability to share free capacity with other logical partitions.
Memory
Depending on given application and performance requirements, a logical partition requests
memory to execute the installed operating system and application. To create Micro-partitions,
the minimum memory requirement is 128 MB per logical partition and dynamically increased
by increments of 16 MB from the overall memory available in the system.
Expansion unit
Expansion units extend the flexibility of the server system to enlarge the number of possible
logical partitions by adding additional hardware, such as storage or network devices.
2.12.6 Specific partition definitions used for Micro-partition
In addition to the base definitions for a Micro-partition, new parameters must be defined to
receive more flexibility and capacity usage of logical partitions included in POWER5
technology.
40p5-550 Technical Overview and Introduction
Capped and uncapped partitions
A capped partition indicates that the local partition will never exceed its assigned capacity. A
partition using the uncapped mode indicates that if the maximum capacity is reached, the
shared pool of processing power is used for more resources if available.
To manipulate the behavior of uncapped partitions, the parameter uncapped weight, in the
range from 0 through 255, must be defined. To prevent an uncapped partition from receiving
extra capacity, the uncapped weight parameter should be 0.
The default uncapped weight is 128.
2.12.7 System Management Service
Either booting up a full partition system or a logical partition to System Management Services
6
(SMS), the ASCII
interface or the GUI are identical in contents and functionality.
The p5-550 (or the logical partition) must be equipped with either a graphic adapter
connected to a graphics display, keyboard, and mouse device, or an ASCII display terminal
connected to one of the native serial ports, or the attached HMC to use the SMS menus. It is
possible to view information about the system (or the single logical partition) and perform
tasks such as setting a password, changing the boot list, and setting the network parameters.
If the system or the partition has been activated without flagging the option to stop to the
SMS, there is the option to press the 1 key on the terminal, or in the graphic window, after the
word keyboard appears and before the word speaker appears. In the terminal, or in the GUI,
the system or the partitions will require you to enter the password defined for admin or
general access. After the text-based SMS starts (either for terminal or graphic window), a
screen similar to the one shown in Figure 2-14 opens.
Version SF220_004
SMS 1.5 (c) Copyright IBM Corp. 2000,2003 All right reserved
----------------------------------------------------------------------------Main Menu
----------------------------------------------------------------------------Type the number of the menu item and press Enter or select Navigation Key:
Figure 2-14 System Management Services main menu
6
American Standard Code for Information Interchange: This is the world-wide standard for the code numbers used
by computers to represent all the uppercase and lowercase Latin letters, numbers, punctuation and so forth.
X = eXit System Management Services
Chapter 2. Architecture and technical overview 41
Note: The version of system firmware currently installed in your system is displayed at the
top of each screen. Processor and other device upgrades might require a specific version
of firmware to be installed in your system.
On each menu screen, you are given the option of choosing a menu item and pressing Enter
(if applicable), or selecting a navigation key. You can use the different options to review or set
the boot list information, or to set up the network environment parameters if you want the
system to boot from a NIM server.
2.12.8 Boot options
The p5-550 handles the boot process in a way that is similar to other pSeries servers.
The initial stage of the boot process is to establish that the machine has powered up correctly
and the memory and CPUs are functioning correctly. After the machine or the logical partition
reaches the SMS menus, all of the necessary tests have been performed and the machine is
scanning the bus for a boot source.
Most system backplanes are designed such that the drive in the first slot spins up
immediately after power-on, and other drives will wait for the operating system to send a
command before spinning up. Disk drive bays 4 and 8 are hardwired to spin-up immediately.
The left-most and accordingly bottom-most slot of the 4-pack disk backplanes (SCSI ID 8,
boot, autostart) is set to spin up immediately after power-on. The power-on delay sequence is
performed to prevent power supply overloading. This behavior makes the disk in the first slot
of the first 4-pack DASD backplane the preferred boot device. See Figure 2-15 to locate all of
the disk bays.
Figure 2-15 Disk bays and SCSI addresses within a p5-550
When SMS menus are available, the Select Boot Options menu can be used to view and set
various options regarding the installation devices and boot devices:
1. Select Install or Boot a Device
Enables you to select a device to boot from or install the operating system from. This
selection is for the current boot only.
42p5-550 Technical Overview and Introduction
2. Select Boot Devices
Enables you to set the boot list.
3. Multiboot Startup
Toggles the multiboot startup flag, which controls whether the multiboot menu is invoked
automatically on startup.
2.12.9 Additional boot options
Instead of booting from the preferred boot device, or from any other internal disks, there are a
number of other possibilities:
DVD-ROM, DVD-RAM These devices can be used to boot the system, or a
Internal or external tape drivesThe media bay tape drive or any externally attached tape
SCSI disk, and Virtual SCSI disk The more common method of booting the system is to
SAN boot It is possible to boot the p5-550 system from a SAN
logical partition (if the resource is available to the specific
partition), so that a system can be loaded, system
maintenance performed, or stand-alone diagnostics
performed.
drive can be used to boot the system, or a logical
partition (if the resource is available to the specific
partition) using mksysb, for example.
use a disk situated in one of the hot-swap bays in the
front of the machine. However, any external
SCSI-attached disk could be used if required. As
described in previous sections, Virtual SCSI devices are
also available to a logical partition.
using a 2 GB Fibre Channel Adapter (FC 6239), or it is
possible to boot one partition using the dedicated 2 GB
Fibre Channel Adapter or the Virtual SCSI device related
to this adapter. The IBM 2105 Enterprise Storage Server
(ESS) is an example of a SAN-attached device that can
provide a boot medium.
LAN boot Network boot and NIM installs can be used if required.
2.12.10 Security
The p5-550 system allows you to set two different types of passwords to limit the access to
these systems. These are defined in the ASMI menus.
system administrator. The
functions and is usually available to all users who are allowed to power on the server,
especially remotely.
2.13 Operating system requirements
All new POWER5 servers are capable of running IBM AIX 5L for POWER and support
appropriate versions of Linux. AIX 5L has been specifically developed and enhanced to
exploit and support the extensive RAS features on IBM Sserver pSeries systems.
Logical partitions can use both a dedicated Ethernet
adapter or Virtual Ethernet to accomplish that.
This password is usually used by the
general ID password provides limited access to the system
Chapter 2. Architecture and technical overview 43
2.13.1 AIX 5L
The p5-550 requires AIX 5L Version 5.3 or AIX 5L Version 5.2 Maintenance Package
5200-04 (IY56722) or later.
The system requires the following media:
AIX 5L for POWER Version 5.2 5765-E62, dated 08/2004 (CD# LCD4-1133-04) or later
AIX 5L for POWER Version 5.3 5765-G03, initial CD set (CD# LCD4-7463-00) or later
IBM periodically releases maintenance packages for the AIX 5L operating system. These
packages are available on CD-ROM (FC 0907), or they can be downloaded from the Internet
at:
http://techsupport.services.ibm.com/server/fixes
You can also get individual operating system fixes and information about obtaining AIX 5L
service at this site. In AIX 5L Version 5.3, there is also the suma command available that helps
the administrator to automate the task of checking and downloading operating system
downloads. For more information about suma command see 3.3.4, “Service Update
Management Assistant” on page 57.
If you have problems downloading the latest maintenance level, ask your IBM Business
Partner or IBM representative for assistance.
The Advanced POWER Virtualization feature is not supported on AIX 5L Version 5.2.
2.13.2 Linux
For the p5-550, Linux distributions are available through SUSE and Red Hat at the time this
publication was written. The p5-550 requires the following version of Linux distributions:
SUSE LINUX Enterprise Server 9 for POWER systems, or later
Red Hat Enterprise Linux AS for POWER Version 3
The Advanced POWER Virtualization feature, DLPAR, and other features require SUSE
SLES 9.
In Japan, Turbolinux is also available. In the Latin America sales region, Conectiva is also
available. For related information and an overview, see:
http://www.ibm.com/servers/eserver/pseries/linux
To find full information about Red Hat Enterprise Linux AS for POWER Version 3 at:
http://www.redhat.com/software/rhel/as/
To find full information about SUSE LINUX Enterprise Server 9 for POWER, see:
IBM only supports the Linux systems of customers with a SupportLine contract covering
Linux. Otherwise, the Linux distributor should be contacted for support.
Chapter 2. Architecture and technical overview 45
46p5-550 Technical Overview and Introduction
3
Chapter 3.Capacity on Demand, RAS, and
manageability
The following sections provide more detailed information about IBM Sserver p5 design
features that will help lower the total cost of ownership (TCO). This chapter includes several
features based on the benefits available when using AIX 5L. Support of these features using
Linux can vary.
The p5-550 systems can be shipped with non-activated processor resources, which can be
added as they are needed. Processors can be brought online to meet increasing workload
demands, without affecting system operations.
The following sections outline the different methods available, namely the processor:
Permanent Capacity Upgrade on Demand
Reserve Capacity on Demand
Dynamic processor sparing
Trial Capacity on Demand for processors and memory
The base 2-way 1.65 GHz POWER5 (FC 5237) processor card supports CUoD. It features
two processors, but zero active, and eight DDR1 Memory DIMM sockets. Processors are
activated in increments of one processor either permanently or for a given amount of time.
Processor Upgrade On Demand and Dynamic Processor Sparing are supported by AIX 5L
Version 5.2 ML4 and AIX 5L Version 5.3. Dynamic Processor Sparing requires that the CPU
guard attribute is set to enable.
3.1.1 Processor Capacity Upgrade on Demand methods
This section describe the different IBM Sserver Capacity Upgrade on Demand methods
available for the processors, at this time of writing.
Permanent Capacity Upgrade on Demand
Capacity Upgrade on Demand (CUoD) is a method where processors are shipped to
customers installed in the p5-550 and later can be activated in increments of one processor.
Additional options deliver the possibility to use temporarily processors resources installed in
the server.
All processor cards are 2-ways, with 0-way active. In an initial order of the p5-550 system, at
least two processors must be activated by ordering the appropriate activation features.
On/Off Capacity on Demand
After an On/Off Processor Enablement feature is ordered and the associated enablement
code is entered into the system, the customer must report On/Off usage to IBM at least
monthly. This information, used to compute the billing data, is then provided to the sales
channel. The sales channel will place an order for a quantity of On/Off Processor Day Billing
features.
Reserve Capacity on Demand
The Reserve Capacity option activation is a reserve capacity for 30 processor-days of prepaid
reserve. To establish processor capacity on the server, select a quantity of inactive
processors to be placed in the server's shared processor pool as reserve processors. After
the server recognizes that non-reserve processors (permanently activated processors)
assigned or available or both to the uncapped partitions have been 100% used , use of
additional processors will cause processor days (good for a 24-hour period) to be subtracted
from the prepaid number of processor-days.
48p5-550 Technical Overview and Introduction
The 30-days Reserve Capacity Processors option is activated with FC 7956. This requires
the purchase of the Advanced POWER Virtualization feature.
Dynamic Processor Sparing
When you have non-activated CUoD processors, a feature called Dynamic Processor Sparing
is automatically provided. Non-activated CUoD processors are processors that are physically
installed in the system, but not activated. Dynamic Processor Sparing allows the system to
disable a failing processor and enable a non-activated CUoD available processor.
Trial Capacity on Demand
Customers with CUoD featured systems must purchase the activation codes from IBM before
the non-activated CUoD resources can be activated to meet the increased workload.
However with the Trial Capacity on Demand feature, customers can activate the required
non-activated CUoD resources immediately and after that, proceed to purchase those
resources from IBM or not. In the HMC, this feature is named
no-cost activation for a maximum period of 30 consecutive days is available as a
complementary service when access to CUoD resources is required immediately.
The following basic rules apply for the p5-550 system:
After the CUoD resources are activated, the customer must either buy part or all of the
activated CUoD resources from IBM or return the activated CUoD resources back to the
system within 30 days.
The Trial CUoD can only be used once.
Activate Immediate. A one-time,
There are several advantages of using this feature:
Improve the response time to meet an unpredictable increase in workload.
The customer can monitor the performance of the system after activating the CUoD
resources before placing the order for activation codes.
It is useful when a CUoD permanent activation purchase is pending.
3.1.2 How to report temporary activation resources
There are three methods for reporting information about the usage of On/Off Capacity on
Demand to IBM (contact yor areas IBM representative for the numbers and addresses):
Using Electronic Service Agent
– Monthly reporting of temporary capacity billing information can be sent to IBM
electronically using the Electronic Service Agent, which is part of the Hardware
Management Console and is designed to monitor events and to transmit server
inventory information to IBM on a periodic, customer-definable timetable.
Using fax
– Fax information about billing information for On/Off Capacity on Demand can be sent
to IBM.
Using e-mail
– E-mail information about billing information for On/Off Capacity on Demand can be
sent to IBM.
Chapter 3. Capacity on Demand, RAS, and manageability 49
3.2 Reliability, availability, and serviceability
Excellent quality and reliability are inherent in all aspects of the IBM Sserver p5 design and
manufacturing. The fundamental objective of the design approach is to minimize outages.
The RAS features help to ensure that the system operates when required, performs reliably,
and efficiently handles any failures that might occur. This is achieved using capabilities
provided by both the hardware and the operating system AIX 5L.
The p5-550 as a POWER5 server enhances the RAS capabilities implemented in
POWER4-based systems. RAS enhancements available on POWER5 servers are:
Most firmware updates allow the system to remain operational.
The ECC has been extended to inter-chip connections for the fabric and processor bus.
Partial L2 cache deallocation is possible.
The number of L3 cache line deletes improved from 2 to 10 for better self-healing
capability.
The following sections describe the concepts that form the basis of leadership RAS features
of IBM Sserver p5 systems in more detail.
3.2.1 Fault avoidance
The p5 systems are built on a quality-based design to keep errors from ever happening. This
design includes the following features:
Reduced power consumption, cooler operating temperatures for increased reliability,
enabled by copper chip circuitry, silicon-on-insulator, and dynamic-clock-gating
Mainframe-inspired components and technologies
3.2.2 First Failure Data Capture
If a problem should occur, the ability to correctly diagnose it is a fundamental requirement
upon which improved availability is based. The p5-550 incorporates advanced capability in
start-up diagnostics and in run-time First Failure Data Capture (FDDC) based on strategic
error checkers built into the chips.
Any errors detected by the pervasive error checkers are captured into Fault Isolation
Registers (FIRs), which can be interrogated by the service processor (SP). The SP in the
p5-550 has the capability to access system components using special purpose service
processor ports or by access to the error registers (Figure 3-1).
50p5-550 Technical Overview and Introduction
Error Checkers
CPU
L1 Cache
L2/L3 Cache
Memor y
F
ault Isolation Register
(unique fingerprint of each
error captured)
Service
Processor
Log Error
Non-volatile
RAM
(FIR)
Disk
Figure 3-1 Schematic of Fault Isolation Register implementation
The FIRs are important because they enable an error to be uniquely identified, thus enabling
the appropriate action to be taken. Appropriate actions might include such things as a bus
retry, ECC correction, or system firmware recovery routines. Recovery routines can include
dynamic deallocation of potentially failing components.
Errors are logged into the system non-volatile random access memory (NVRAM) and the SP
event history log, along with a notification of the event to AIX for capture in the operating
system error log. Diagnostic Error Log Analysis (
diagela) routines analyze the error log
entries and invoke a suitable action such as issuing a warning message. If the error can be
recovered, or after suitable maintenance, the service processor resets the FIRs so that they
can accurately record any future errors.
The ability to correctly diagnose any pending or firm errors is a key requirement before any
dynamic or persistent component deallocation or any other reconfiguration can take place.
For more information, see “Dynamic or persistent deallocation” on page 53.
3.2.3 Permanent monitoring
The SP included in the p5-550 provides a means to monitor the system even when the main
processor is inoperable. See the following subsections for a more detailed description of
monitoring functions in p5-550.
Mutual surveillance
The SP can monitor the operation of the firmware during the boot process, and it can monitor
the operating system for loss of control. This allows the service processor to take appropriate
action, including calling for service, when it detects that the firmware or the operating system
has lost control. Mutual surveillance also allows the operating system to monitor for service
processor activity and can request a service processor repair action if necessary.
Environmental monitoring
Environmental monitoring related to power, fans, and temperature is done by the System
Power Control Network (SPCN). Environmental critical and non-critical conditions generate
Early Power-Off Warning (EPOW) events. Critical events (for example, Class 5 AC power
loss) trigger appropriate signals from hardware to impacted components so as to prevent any
data loss without the operating system or firmware involvement. Non-critical environmental
events are logged and reported through Event Scan.
Chapter 3. Capacity on Demand, RAS, and manageability 51
The operating system cannot program or access the temperature threshold using the SP.
EPOW events can, for example, trigger the following actions:
Temperature monitoring, which increases the fans speed rotation when ambient
temperature is above a preset operating range.
Temperature monitoring warns the system administrator of potential
environmental-related problems. It also performs an orderly system shutdown when the
operating temperature exceeds a critical level.
Voltage monitoring provides warning and an orderly system shutdown when the voltage is
out of the operational specification.
3.2.4 Self-healing
For a system to be self-healing, it must be able to recover from a failing component by first
detecting and isolating the failed component, taking it off line, fixing or isolating it, and
reintroducing the fixed or replacement component into service without any application
disruption. Examples include:
Bit steering to redundant memory in the event of a failed memory module to keep the
server operational.
Bit-scattering, thus allowing for error correction and continued operation in the presence
of a complete chip failure (
Single bit error correction using ECC without reaching error thresholds for main, L2, and
L3 cache memory.
Chipkill recovery).
L3 cache line deletes extended from 2 to 10 for additional self-healing.
ECC extended to inter-chip connections on fabric and processor bus.
Memory scrubbing to help prevent soft-error memory faults.
Dynamic processor deallocation, a deallocated processor can be replaced by an unused
Capacity on Demand processor to keep the system operational.
Memory reliability, fault tolerance, and integrity
The p5-550 uses Error Checking and Correcting (ECC) circuitry for system memory to correct
single-bit and to detect double-bit memory failures. Detection of double-bit memory failures
helps maintain data integrity. Furthermore, the memory chips are organized such that the
failure of any specific memory module only affects a single bit within a four-bit ECC word
bit-scattering), thus allowing for error correction and continued operation in the presence of
(
a complete chip failure (
and thresholding to determine when spare memory modules within each bank of memory
should be used to replace ones that have exceeded their threshold of error count (
bit-steering
idle time and checking and correcting any single-bit errors that have accumulated by passing
the data through the ECC logic. This function is a hardware function on the memory controller
chip and does not influence normal system memory performance.
). Memory scrubbing is the process of reading the contents of the memory during
3.2.5 N+1 redundancy
The use of redundant parts allows the p5-550 to remain operational with full resources:
Chipkill recovery). The memory DIMMs also use memory scrubbing
dynamic
Redundant spare memory bits in L1, L2, L3, and main memory
Redundant fans
Redundant power supplies (optional)
52p5-550 Technical Overview and Introduction
3.2.6 Fault masking
If corrections and retries succeed and do not exceed threshold limits, the system remains
operational with full resources, and no client or IBM customer engineer intervention is
required. This technology is used in the following faults:
CEC bus retry and recovery
PCI-X bus recovery
ECC Chipkill soft error
3.2.7 Resource deallocation
If recoverable errors exceed threshold limits, resources can be deallocated with system
remaining operational, allowing deferred maintenance at a convenient time.
Dynamic or persistent deallocation
Dynamic deallocation of potentially failing components is non-disruptive, allowing the system
to continue to run. Persistent deallocation occurs when a failed component is detected, which
is then deactivated at a subsequent reboot.
Dynamic deallocation functions include:
Processor
L3 cache line delete
Partial L2 cache deallocation
PCI-X bus and slots
For dynamic processor deallocation, the service processor performs a predictive failure
analysis based on any recoverable processor errors that have been recorded. If these
transient errors exceed a defined threshold, the event is logged and the processor is
deallocated from the system while the operating system continues to run. This feature
(named
deallocation can only occur if there are sufficient functional processors (at least two).
To verify whether CPU Guard has been enabled, run the following command:
lsattr -El sys0 | grep cpuguard
If enabled, the output will be similar to the following:
cpuguard enable CPU Guard True
If the output shows CPU Guard as disabled, enter the following command to enable it:
chdev -l sys0 -a cpuguard='enable'
Cache or cache-line deallocation is aimed at performing dynamic reconfiguration to bypass
potentially failing components. This capability is provided for both L2 and L3 caches. Dynamic
run-time deconfiguration is provided if a threshold of L1 or L2 recovered errors is exceeded.
In the case of an L3 cache run-time array single-bit solid error, the spare chip resources are
used to perform a line delete on the failing line.
CPU Guard) enables maintenance to be deferred until a suitable time. Processor
PCI hot-plug slot fault tracking helps prevent slot errors from causing a system machine
check interrupt and subsequent reboot. This provides superior fault isolation, and the error
affects only the single adapter. Run-time errors on the PCI bus caused by failing adapters will
result in recovery action. If this is unsuccessful, the PCI device will be gracefully shut down.
Chapter 3. Capacity on Demand, RAS, and manageability 53
Parity errors on the PCI bus itself will result in bus retry, and if uncorrected, the bus and any
I/O adapters or devices on that bus will be deconfigured.
The p5-550 supports PCI Extended Error Handling (EEH) if it is supported by the PCI-X
adapter. In the past, PCI bus parity errors caused a global machine check interrupt, which
eventually required a system reboot in order to continue. In the p5-550 system, hardware,
system firmware, and AIX interaction have been designed to allow transparent recovery of
intermittent PCI bus parity errors and graceful transition to the I/O device available state in the
case of a permanent parity error in the PCI bus.
EEH-enabled adapters respond to a special data packet generated from the affected PCI slot
hardware by calling system firmware, which will examine the affected bus, allow the device
driver to reset it, and continue without a system reboot.
Following a hardware error that has been flagged by the service processor, the subsequent
reboot of the system will invoke extended diagnostics. If a processor or L3 cache has been
marked for deconfiguration by persistent processor deallocation, the boot process will attempt
to proceed to completion with the faulty device automatically deconfigured. Failing I/O
adapters will be deconfigured or bypassed during the boot process.
Note: The auto-restart (reboot) option, when enabled, can reboot the system automatically
following an unrecoverable software error, software hang, hardware failure, or
environmentally induced failure (such as loss of power supply)
3.2.8 Serviceability
Increasing service productivity means the system is up and running for a longer time. p5-550
improves service productivity by providing the functions described in the following
subsections:
Error indication and LED indicators
The p5-550 is designed for customer setup of the machine and for the subsequent addition of
most hardware features. The p5-550 also allows customers to replace service parts
(Customer Replaceable Unit). To accomplish this, the p5-550 provides internal LED
diagnostics that will identify parts that require service. Attenuation of the error is provided
through a series of light attention signals, starting on the exterior of the system (System
Attention LED) located on the front of the system, and ending with an LED near the failing
Field Replaceable Unit.
For more information about Customer Replaceable Units, including videos, see:
http://publib.boulder.ibm.com/eserver
System Attention LED
The attention indicator is represented externally by an amber LED on the operator panel and
the back of the system unit. It is used to indicate that the system is in one of the following
states:
Normal state, LED is off.
Fault state, LED is on solid.
54p5-550 Technical Overview and Introduction
Identify state, LED is blinking.
Additional LEDs on I/O components such as PCI-X slots and disk drives provide status
information such as power, hot-swap, and need for service.
Concurrent Maintenance
Concurrent Maintenance provides replacement of the following parts while the system
remains running:
Disk drives
Cooling fans
Power subsystems
PCI-X adapter cards
3.3 Manageability
The functions and tools provided for IBM Sserver p5 systems are described in the next
sections.
3.3.1 Service processor
With system in power standby mode, or with an operating system in control of the machine, or
controlling the related partition, the SP is working and checking the system for errors,
ensuring the connection to the HMC for manageability purposes. With system up and
running, the SP provides the possibility to view and change the Power-On settings, using the
Advanced System Management Interface (ASMI). Also, the surveillance function of the SP is
monitoring the operating system to check that it is still running and has not stalled.
See Figure 3-2 for an example of the Advanced System Management Interface accessed
from a Web browser.
Chapter 3. Capacity on Demand, RAS, and manageability 55
Figure 3-2 Advanced System Management main menu
3.3.2 Service Agent
Service Agent is an application program that operates on a p5, pSeries or IBM RS/6000
computer and monitors them for hardware errors. It reports detected errors, assuming they
meet certain criteria for criticality, to IBM for service with no customer intervention. It is an
enhanced version of Service Director with a graphical user interface.
Key things you can accomplish using Service Agent for p5, pSeries, and RS/6000 include:
Automatic problem analysis
Problem-definable threshold levels for error reporting
Automatic problem reporting; service calls placed to IBM without intervention
Automatic customer notification
In addition:
Commonly viewed hardware errors. You can view hardware event logs for any monitored
machine on the network from any Service Agent host user interface.
High-availability cluster multiprocessing (HACMP) support for full fallback. Includes
high-availability cluster workstation (HACWS) for 9076.
Network environment support with minimum telephone lines for modems.
56p5-550 Technical Overview and Introduction
VPD data can be sent to IBM using Performance Management.
Machines are defined by using the Service Agent user interface. After the machines are
defined, they are registered with the IBM Service Agent Server (SAS). During the registration
process, an electronic key is created that becomes part of your resident Service Agent
program. This key is used each time the Service Agent places a call for service. The IBM
Service Agent Server checks the current customer service status from the IBM entitlement
database; if this reveals that you are not on Warranty or MA, the service call is refused and
posted back through an e-mail notification.
Service Focal Point
Service Focal Point is used by service technicians to start and end their service calls. It
provides service representatives with event, Vital Product Data (VPD), and diagnostic
information. The HMC can also notify service representatives of hardware failures
automatically by using the Service Agent features. You can configure the HMC to use the
Service Agent call-home feature to send IBM event information. This information is stored,
analyzed, and then acted upon by the service representative. Some parts of Service Focal
Point need to be configured so that the proper information is sent to IBM.
You can download the latest version of Service Agent at:
ftp://ftp.software.ibm.com/aix/service_agent_code
3.3.3 p5 Customer-Managed Microcode
The pSeries and RS/6000 Customer-Managed Microcode is a methodology that enables you
to manage and install microcode updates on p5, pSeries and RS/6000 systems and
associated I/O adapters. The IBM pSeries Microcode Update Web site can be found at:
IBM provides service tools that can assist you in determining microcode levels and updating
systems with the latest available microcode. To determine which tool to use in a specific
environment, visit:
The Service Update Management Assistant (SUMA) helps system administrators retrieve
maintenance updates from the Web. SUMA offers flexible options that let customers set up
policies to automate the download of fixes to their systems. SUMA policies can be scheduled
to periodically check the availability of specific new fixes (APAR, PTF, or fileset), critical or
security fixes, or an entire maintenance level. A notification e-mail can be sent detailing
updates that are needed when comparing available fixes to installed software, a fix
repository, or a maintenance level.
Benefits provided by SUMA:
Moves administrators away from the task of manually retrieving maintenance updates
from the Web.
Policy can be scheduled to run periodically, for example, to download the latest critical
fixes weekly.
Can compare fixes needed against software inventory, fix repository, or a maintenance
level.
Receive mail notification after a fileset preview or download operation.
Chapter 3. Capacity on Demand, RAS, and manageability 57
Allows for FTP, HTTP, or secure HTTPS transfers.
Provides the same requisite checking as the IBM fix distribution Web site.
Available through SMIT menus (smitty suma) or a command line interface.
3.4 Cluster 1600
Today's IT infrastructure requires that systems meet increasing demands, while offering the
flexibility and manageability to rapidly develop and deploy new services. IBM clustering
hardware and software provide the server, with availability, scalability, security, and
single-point-of-management control, to satisfy these needs.
IBM Sserver Cluster 1600 is a POWER-based AIX 5L and Linux Cluster targeting scientific
and technical computing, large-scale databases, and workload consolidation
IBM Cluster Systems Management (CSM) is designed to provide a robust, powerful, and
centralized way to manage a large number of POWER5-based systems all from one single
point of control. CSM can help lower the overall cost of IT ownership by helping to simplify the
tasks of installing, operating, and maintaining clusters of servers. CSM can provide one
consistent interface for managing both AIX and Linux nodes (physical systems or logical
partitions), with capabilities for remote parallel network install, remote hardware control, and
distributed command execution.
The p5-550 is supported with the Cluster 1600 running CSM for AIX, V1.3.1. To attach a
p5-550 to a Cluster 1600, an HMC is required. One HMC can also control several p5-550s
that are part of the cluster. If a p5-550 configured in partition mode (with physical or virtual
resources) is part of the cluster, all partitions must be part of the cluster.
It is not possible to use selected partitions as part of the cluster and use others for non-cluster
use. The HMC uses a dedicated connection to the p5-550 to provide the functions needed to
control the server, such as powering the system on and off. The HMC must have an Ethernet
connection to the Control Work Station (CWS). Each partition in p5-550 must have an
Ethernet adapter to connect to the CWS
Information regarding HMC control, cluster building block servers, and cluster software
available at:
http://www.ibm.com/servers/eserver/clusters/
trusted LAN.
The benefits of a clustered environment based on logical partitions
The evolution of processor and storage technologies has a great impact on the architecture of
IT infrastructures. This was the most significant challenge for the infrastructure in the past
and will also be in the future. During the first half of the 1990s, one single central instance of
an application per node was suitable, moreover, most productive systems needed
additionally associated nodes, so-called application servers.
Increasing performance and reliability by simply replicating application server nodes led to
complex environments that often resulted in poor system management. The reason for these
complex constructions was the limited computing power of a single node. This limitation was
softened during the second half of the 1990s.
Big symmetric multiprocessor (SMP) nodes with higher clock rates and increased memory
provided the possibility to install more than one system on a node. This had some side effects
regarding systems operations: Release planning processes had to pay attention to different
databases or application versions, or both, to avoid unresolved conflicts.
58p5-550 Technical Overview and Introduction
In 2000, Workload Manager for AIX (WLM) was announced. Multiple application instance
installations became more and more popular because of the permanently increasing number
of systems dedicated to applications at customer sites. The general availability of this
functionality of AIX to separate the workloads of dedicated systems eliminated the last
obstacle for consolidating several systems in one node.
Some customers expanded the usage of their dedicated systems and consequently model
more business processes. This often caused an increased number of dedicated systems
used and a stronger demand on flexibility. In addition, the life cycle of these systems differed
extremely. Renaming, removal, or deletion became more and more common system
administration tasks.
In 2001, the pSeries hardware technology with logical partitioning was generally available.
Logical partitioning creates the possibility to define the logical partitions (LPARs) that are
adapted to customer needs regarding the number of processors, assigned memory, and I/O
adapters. The p5-550 offers the flexibility to increase the usage of the resources even more,
and reduce the total cost of ownership (TCO).
Partitions with associated physical resources or virtual resources are not different from a
collection of stand-alone nodes.
Today, server consolidation is a must for many IT sites. Minimized TCO and complexity, with
the maximum amount of flexibility, is a crucial goal of nearly all customers. LPARs allow a
flexible distribution of resources with LPAR boundaries. Each logical partition can be
configured according to the specific needs of the occupant application. LPARs provide a
protection boundary between the systems. More test and development systems can exist on
the same server in separate partitions.
The CSM value points
The CSM allows the management of different hardware platforms from one single point of
control and it has consistent interfaces to manage systems and logical partitions running both
AIX and Linux. The management is achieved across multiple switch and interconnect
topologies. PSSP forced system administrators to do some things a certain way (such as
NIM, and SP user management). The CSM provides assistance on setting these things up,
but allows the system administrator to tailor their system to their own needs, and it has the
ability to manage systems across different geographical sites.
Monitoring is much easier to use and the system administrator can monitor all the network
interfaces, not just the switch and administrative interfaces. The management server pushes
information out to the nodes, which allows the management server to not have to trust the
node. In addition, the nodes do not have to be network connected to each other either. This
means that giving root access on one node does not mean giving root access on all the
nodes. The base security setup is all done automatically at install time.
The CSM ships with AIX itself (a 60-day Try and Buy license is shipped with AIX). The CSM
client side is automatically installed and ready when you install AIX, so each system or logical
partition is cluster-ready.
CSM Version 1.4 on AIX and Linux (planned 4Q04)
The CSM V1.4 on AIX and Linux introduces an optional IBM CSM High Availability
Management Server feature, designed to allow automated failover of the CSM management
server to a backup management server. In addition, sample scripts for setting up NTP
1
, and
1
Network Time Protocol
Chapter 3. Capacity on Demand, RAS, and manageability 59
network tuning (AIX ONLY) configurations, and the capability to copy files across nodes or
node groups in the cluster can improve cluster ease of use and site customization.
60p5-550 Technical Overview and Introduction
Related publications
The publications listed in this section are considered particularly suitable for a more detailed
discussion of the topics covered in this Redpaper.
IBM Redbooks
For information about ordering these publications, see “How to get IBM Redbooks” on
page 63. Note that some of the documents referenced here may be available in softcopy only.
BM Sserver pSeries 670 and pSeries 690 System Handbook, SG24-7040
The Complete Partitioning Guide for IBM pSeries Servers, SG24-7039
Managing AIX Server Farms, SG24-6606
Practical Guide for SAN with pSeries, SG24-6050
Problem Solving and Troubleshooting in AIX 5L, SG24-5496
Understanding IBM Sserver pSeries Performance and Sizing, SG24-4810
Other publications
These publications are also relevant as further information sources:
7014 Series Model T00 and T42 Rack Installation and Service Guide, SA38-0577,
contains information regarding the 7014 Model T00 and T42 Rack, in which this server
can be installed.
7316-TF3 17-Inch Flat Panel Rack-Mounted Monitor and Keyboard Installation and
Maintenance Guide, SA38-0643, contains information regarding the 7316-TF3 Flat Panel
Display, which can be installed in your rack to manage your system units.
IBM Sserver Hardware Management Console for pSeries Installation and Operations
Guide, SA38-0590, provides information to operators and system administrators on how
to use a IBM Hardware Management Console for pSeries (HMC) to manage a system. It
also discusses the issues associated with logical partitioning planning and
implementation.
Planning for Partitioned-System Operations, SA38-0626, provides information to
planners, system administrators, and operators about how to plan for installing and using
a partitioned server. It also discusses some issues associated with the planning and
implementing of partitioning.
RS/6000 and Sserver pSeries Adapters, Devices, and Cable Information for Multiple Bus
Systems, SA38-0516, contains information about adapters, devices, and cables for your
system. This manual is intended to supplement the service information found in the
Diagnostic Information for Multiple Bus Systems documentation.
RS/6000 and Sserver pSeries Diagnostics Information for Multiple Bus Systems,
SA38-0509, contains diagnostic information, service request numbers (SRNs), and failing
function codes (FFCs).
RS/6000 and pSeries PCI Adapter Placement Reference, SA38-0538, contains
information regarding slot restrictions for adapters that can be used in this system.
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
Help from IBM
IBM Support and downloads
ibm.com/support
IBM Global Services
ibm.com/services
Related publications 63
64p5-550 Technical Overview and Introduction
Back cover
IBM Eserver p5 550
Technical Overview
®
and Introduction
Finer system
granularity using
Micro-Partitioning
technology to lower
TCO
Outstanding
performance based
on POWER5 processor
technology
From Web servers to
integrated cluster
solutions
This document is a comprehensive guide covering the IBM
Sserver p5 550 UNIX servers. We introduce major hardware
offerings and discuss their prominent functions.
Professionals wishing to acquire a better understanding of IBM
Sserver p5 products should consider reading this document.
The intended audience includes:
Customers
Sales and marketing professionals
Technical support professionals
IBM Business Partners
Independent software vendors
This document expands the current set of IBM
documentation by providing a desktop reference that offers a
detailed technical description of the p5-550 system.
This publication does not replace the latest pSeries marketing
materials and tools. It is intended as an additional source of
information that, together with existing sources, can be used to
enhance your knowledge of IBM server solutions.
Sserver
Redpaper
INTERNATIONAL
TECHNICAL
SUPPORT
ORGANIZATION
BUILDING TECHNICAL
INFORMATION BASED ON
PRACTICAL EXPERIENCE
IBM Redbooks are developed
by the IBM International
Technical Support
Organization. Experts from
IBM, Customers and Partners
from around the world create
timely technical information
based on realistic scenarios.
Specific recommendations
are provided to help you
implement IT solutions more
effectively in your
environment.
For more information:
ibm.com/redbooks
Loading...
+ hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.