Академический Документы
Профессиональный Документы
Культура Документы
Table of Contents
0. Document Control.............................................................................................................................................................6
0.1. Revision History.......................................................................................................................................................6
0.2. Document References.............................................................................................................................................6
1. Introduction.......................................................................................................................................................................7
2. Overview...........................................................................................................................................................................8
2.1. Product Line.............................................................................................................................................................8
2.2. Meet the processors................................................................................................................................................9
2.2.1. Olympus Processor.........................................................................................................................................9
2.2.2. Jupiter Processor............................................................................................................................................9
2.3. Meet the Servers....................................................................................................................................................10
2.4. Solaris Support......................................................................................................................................................10
3. Servers...........................................................................................................................................................................11
3.1. Sun SPARC Enterprise M3000 Server..................................................................................................................11
3.1.1. M3000 Overview...........................................................................................................................................11
3.1.2. M3000 Broad Specs......................................................................................................................................11
3.1.3. M3000 Front View.........................................................................................................................................12
3.1.4. M3000 Rear View..........................................................................................................................................12
3.1.5. M3000 Top View...........................................................................................................................................13
3.2. Sun SPARC Enterprise M4000 Server..................................................................................................................14
3.2.1. M4000 Overview...........................................................................................................................................14
3.2.2. M4000 Broad Specs......................................................................................................................................14
3.2.3. M4000 Front View.........................................................................................................................................15
3.2.4. M4000 Rear View..........................................................................................................................................15
3.2.5. M4000 Top View...........................................................................................................................................16
3.3. Sun SPARC Enterprise M5000 Server..................................................................................................................17
3.3.1. M5000 Overview...........................................................................................................................................17
3.3.2. M5000 Broad Specs......................................................................................................................................17
3.3.3. M5000 Front View.........................................................................................................................................18
3.3.4. M5000 Rear View..........................................................................................................................................18
3.3.5. M5000 Top View...........................................................................................................................................19
3.4. Sun SPARC Enterprise M8000 Server..................................................................................................................20
3.4.1. M8000 Overview...........................................................................................................................................20
3.4.2. M8000 Broad Specs......................................................................................................................................20
3.4.3. M8000 Front View.........................................................................................................................................21
3.4.4. M8000 Rear View..........................................................................................................................................22
3.5. Sun SPARC Enterprise M9000/M9000+ Server....................................................................................................23
3.5.1. M9000/M9000+ Overview.............................................................................................................................23
3.5.2. M9000 Expansion Cabinet (M9000+)............................................................................................................23
3.5.3. M9000-32 Broad Specs................................................................................................................................24
3.5.4. M9000-64 Broad Specs................................................................................................................................24
3.5.5. M9000 Front View.........................................................................................................................................25
3.5.6. M9000 Rear View..........................................................................................................................................26
3.5.7. M9000+ Front View.......................................................................................................................................27
3.5.8. M9000+ Front View (No Cables)...................................................................................................................27
3.5.9. M9000+ Rear View........................................................................................................................................28
4. Components...................................................................................................................................................................29
4.1. Hot Plug.................................................................................................................................................................30
4.2. Hot Swap...............................................................................................................................................................30
4.3. PSU........................................................................................................................................................................30
4.4. FANU.....................................................................................................................................................................31
4.4.1. Form Factor Servers.....................................................................................................................................31
4.4.2. Data Center Servers.....................................................................................................................................31
4.4.3. Air Flow.........................................................................................................................................................31
4.5. MBU.......................................................................................................................................................................32
4.6. Backplane..............................................................................................................................................................33
4.6.1. BP_A.............................................................................................................................................................33
4.6.2. BP_B.............................................................................................................................................................33
4.7. XBU........................................................................................................................................................................33
4.8. CLKU.....................................................................................................................................................................33
4.9. CPU & Memory......................................................................................................................................................34
4.9.1. Memory Subystem........................................................................................................................................35
4.9.1.1. SC.........................................................................................................................................................35
4.9.1.2. MAC......................................................................................................................................................35
4.9.1.3. DIMM's..................................................................................................................................................36
0. Document Control
0.1. Revision History
Version Author Reason For Issue Date
1.0 Kevin Hickman Draft for Review 12/02/09
2.0 Kevin Hickman Update after Review 11/08/09
Removal of Client specific references
1. Introduction
This document is Sun Confidential and is only to be used by Sun badged engineers who meet the pre-requisites
listed below.
This guide has been written to introduce the concepts of, and to demonstrate basic operation and administration
functions of the five joint Sun-Fujitsu OPL (Oympus Product Line) Servers, namely the M3000, M4000, M5000, M8000
and M9000 SPARC Enterprise Servers.
Based on the dual-core SPARC64 VI processor, the quad-core SPARC64 VII processor and the Solaris Operating
System (OS), the OPL (Olympus Product Line) SPARC Enterprise servers are ideal platforms for high-availability,
mission-critical enterprise applications including large-scale databases, BIDW, ERP and CRM.
The guide aims to provide the user with enough information to meet the following objectives:
• Describe the architecture, functional goals, and reliability, availability, and serviceability (RAS) features of the
Sun SPARC Enterprise MX000 servers
• Identify system components by location, layout, and function
• Perform configuration and platform administration methodologies supported by Sun on the Sun SPARC
Enterprise MX000 servers
• Be familiar with and use troubleshooting tools and techniques available within the Solaris OS
• Configure and manage the platform using service processor (XSCF) utilities
• Initialize and configure the Solaris OS on a domain
• Define and edit nonvolatile random access memory (NVRAM) parameters
• Perform flash programmable read-only memory (PROM) image updates
• Create, manage, and delete multiple domains using hardware, system configuration, and domain component
lists (DCLs)
• Use OpenBoot programmable read-only memory (PROM) commands to gather system information and
interpret results
To make the most use out of the guide, the user should already be able to perform the following functions without
assistance:
2. Overview
2.1. Product Line
The MX000 servers are the result of an alliance between Sun Microsystems and Fujitsu. These servers are Sun
and Fujitsu’s mid-range and high-end server models which have been jointly designed and manufactured by
Sun and Fujitsu. They run the Solaris 10 OS and are SPARC v9-compliant for binary compatibility with SPARC-
based Sun Fire™ and Fujitsu PRIMEPOWER servers and are fully compliant with Sun’s standards for software,
storage, and services. This alliance entails shared investment of MX000 servers’ product development costs
and has allowed Sun to invest more heavily in chip multithreading (CMT) technology for future enterprise
computing products.
• FF (Form Factor) is the code name for the systems from the Olympus Product Line. The FF product
family comprises 6U (FF1/M4000) and 10U (FF2/M5000) rack optimized servers with dual core
SPARC64 VI processors or quad core SPARC64 VII processors.
• The SPARC Enterprise M8000 and M9000 comprise the DC (Data Center) product line. Together, FF
and DC servers form the Olympus Product Line (OPL). OPL and LE (SPARC Enterprise T1000 &
T2000) form the Advanced Product Line (APL)
Note:
Sun SPARC Enterprise M4000, M5000, M8000, and M9000 servers utilize SPARC64 VI and SPARC64 VII
processors developed by Fujitsu. The SPARC64 VI dual-core, multithreaded processor takes advantage of
90nm technology while the SPARC64 VII processor provides a quad-core implementation with a faster clock
speed and a reduction in size using 65 nm fabrication. In all Servers, except the M3000, these CPU's can be
mixed.
For reference, there are several features of the Fujitsu SPARC VI+ Processors which you may come across in
the documentation:
• Chip Multi Processing (CMP), which provides multiple cores per CPU
• Chip Multi Threading (CMT), which provides multiple threads per core
• Vertical Multi Threading (VMT) technology, which allows multiple threads to run in parallel
Note
The CPUs have two physical cores and each core has two strands with VMT structures. This allows four
threads to run in parallel. The two strands that belong to the same physical core share most of the
physical resources, while the two physical cores do not share physical resources except the Level 2 (L2)
cache and system interface.
The SPARC64 VI processor consists of Dual Core Processor with 2 Vertical Threads Per Core. It has a
Frequency Range of 2.15-2.4GHz at Launch and Instruction and Data Caches of 128KB each Per Core.
There is 5/6MB On Chip L2 Cache Per Chip and it supports and Out of Order Execution capability.
The new SPARC64 VII processor consists of four SPARC V9 cores with two simultaneous threads per
core. It's built with an advanced 65nm Cu 10-layer process technology for high speed and density, and is
100% binary compatible with the SPARC64 VI process so there is no need to recompile.
All OPL systems support the Jupiter processor. The OPL systems also support Mixed CPU speeds, but
the XSCF require firmware version 1070 or higher to run Jupiter.
Server Max No. Max. No. Max No. Max No. Of Max No. of Max No. Max No.
of of of RAM Internal External of I/O of
Procs Cores Threads (GB) I/O Slots I/O Units Slots Domains
M3000 1 4 8 32 4 Not Supported 4 1
M4000 4 16 32 128 5 2 25 2
M5000 8 32 64 256 10 4 50 4
M8000 16 64 128 512 32 8 112 16
M9000-32 32 128 256 1024 64 16 224 24
M9000-64 64 256 512 4096 128 16 288 24
Notes:
• Maximum RAM capacity for Form Factor Servers is calculated based on 4-Gbyte DIMM
• Maximum RAM capacity for Data Center Servers is calculated based on 8-Gbyte DIMM
• Maximum CPU statistics for all Servers is based on the quad-core SPARC64 VII CPU (Jupiter)
In Summary, including the Entry Level M3000, there are 5 Servers in the OPL Product Range. For reference, the
next generation of Sun Servers will be project SuperNova which uses the “Rock” CPU which is due in July 2009.
Notes:
• The M3000 only supports SPARC64 VII CPU's and therefore requires Solaris 10_u4 (8/07) or higher
• In order to support COD and DR you will need Solaris 10_u4 8/07. In the M3000 server, the COD
function is not available.
3. Servers
3.1. Sun SPARC Enterprise M3000 Server
3.1.1. M3000 Overview
The SPARC Enterprise M3000 (code named Ikkaku.) is a single socket (single processor ) SPARC64 VII
based 2RU server which offers extreme RAS in a 2RU footprint. Hot swappable power supplies, fans and
disk drives along with XSCF, automatic recovery and instruction level retry, Extended ECC memory
protection, and Fault Isolation (Chip,processors, I/O, ASICS) rich check/recovery mechanisms in the
processor eliminate planned or unplanned downtime.
The SPARC Enterprise M3000 server maximum configuration is equipped with the 2.52 GHz SPARC64
VII quad-core processor, up to four 146 GB SAS disks, 32 GB of memory, and one, two-lane SAS port for
additional storage connectivity. Additionally, the M3000 is outfitted with four low profile PCIe slots, one
RCI, one USB, one Serial, two LAN 0 & 1 connections, two UPC, and four Gigabit Ethernet slots.
The advanced design of the Sun SPARC Enterprise M3000 server architecture reduces the number of
working parts eliminating multiple points of failure. And because it is a SPARC64 based server, it can be
upgraded to the next generation processor within the same box.
• 1 Domain
• 1 XSCFU (eXtended System Control Facility Unit) Service Processor
• 2 RU Server (22kg)
• 1 CPU - SPARC64 VII quad-core 2.52GHz, 4 cores and 8 threads
• 32GB RAM (8 DIMM Slots using 4GB DIMMs)
• 4 PCIe Low Profile slots, no hot plug
• 4 x 2.5" (146gb) SAS hot swap HDD
• 1 Internal DVD Drive
• 4 On board GBE
• 2 PSU (n + 1), optional dual feed, hot swap
• n + 1 Cooling Fan, hot swap
Note that the OPL product line (M4000 - M9000) only supports Solaris 10_u3 11/06 and above. In order to
support COD and DR you will need Solaris 10_u4 8/07
Built for data center RAS demands: End-to-end ECC protection, ECC and Extended ECC protection for
memory, memory mirroring, and predictive self healing, guaranteed data path protection, and more. An
all-new, faster backplane interconnect gives you 32 GB/second speeds and acts as a common
interconnect between all chips, maximizing address and data bandwidth on one interconnect.
• 2 Domains
• 1 XSCFU (eXtended System Control Facility Unit) Service Processor
• 6 RU server (84kg)
• 2 CPUM (CPU Modules – 2 CPU sockets per board)
• 4 CPUs - SPARC64 VI dual-core and/or SPARC64 VII quad-core (2.1 or 2.4 GHz),
therefore up to 16cores and 32 threads
• 4 MEMB (Memory Boards)
• 128GB RAM (32 DIMM Slots using 4GB DIMMs)
• 1 IOU (5 cassettes)
• 4 PCIe slots and 1 PCI-X
• Up to 2 External I/O Expansion Units
• Up to 25 PCIe or PCI-X slots with optional I/O Expansion Units
• 1 CD-RW/DVD-RW drive unit,
• 2 Internal 2.5" (146GB) SAS HDDs
• 1 tape drive unit (optional)
• 2 built in Gigabit Ethernet ports
• 2 PSU (n + 1) Hot-Swap Single Phase
• 4 fans per server for cooling.
• Two 172-mm fans (FAN_A). One fan is redundant
• Two 60-mm fans (FAN_B). One fan is redundant)
Note that the OPL product line (M4000 - M9000) only supports Solaris 10_u3 11/06 and above. In order to
support COD and DR you will need Solaris 10_u4 8/07
Built for datacenter RAS demands: End-to-end ECC protection, ECC and Extended ECC protection for
memory, memory mirroring, and predictive self healing, guaranteed data path protection, and more. An
all-new, faster backplane interconnect gives you 64 GB/second speeds and acts as a common
interconnect between all chips, maximizing address and data bandwidth on one interconnect.
• 4 Domains
• 1 XSCFU (eXtended System Control Facility Unit) Service Processor
• 10 RU server (125kg)
• 4 CPUM (CPU Modules – 2 CPU sockets per board)
• 8 CPUs - SPARC64 VI dual-core and/or SPARC64 VII quad-core (2.1 or 2.4 GHz),
therefore up to 32 cores and 64 threads
• 8 MEMB (Memory Boards)
• Up to 256 GB RAM (64 DIMM Slots using 4GB DIMMs)
• 2 IOU (10 cassettes)
• 8 PCIe slots and 2 PCI-X
• Up to 4 External I/O Expansion Units
• Up to 50 PCIe or PCI-X slots with optional I/O Expansion Units
• 1 CD-RW/DVD-RW drive unit,
• 4 Internal 2.5" (146GB) SAS HDDs
• 1 tape drive unit (optional)
• 4 built in Gigabit Ethernet ports
• 4 PSU (n + 2) Hot-Swap Single Phase
• Four 172-mm (FAN_A) fans per server for cooling. Two fans are redundant
The Sun SPARC Enterprise M8000 servers can support up to 16 dynamic domains. For pure flexibility,
we extended the domain partitioning granularity down to one processor chip, four DIMMs and two I/O
slots.
Note that the OPL product line (M4000 - M9000) only supports Solaris 10_u3 11/06 and above. In order to
support COD and DR you will need Solaris 10_u4 8/07
The Sun SPARC Enterprise M8000 server features the option to mix and match up to 16 powerful quad-
core SPARC64 VII processors running at 2.4-2.52 GHz or 16 dual-core SPARC64 VI processors and
comes preinstalled with Solaris 10 OS, providing binary compatibility and maximum investment
protection.
Built for data center RAS demands: End-to-end ECC protection, ECC and Extended ECC protection for
memory, memory mirroring, and predictive self healing, guaranteed data path protection, and more. An
all-new, faster backplane interconnect gives you 184 GB/second speeds and acts as a common
interconnect between all chips, maximizing address and data bandwidth on one interconnect.
The I/O system interface features hot-swappable I/O Units (IOUs), each of which provides eight PCIe
slots and four 2.5-inch SAS disk drive bays. Fully configured, the server supports up to 16 internal SAS
boot disks, 32 internal PCIe slots, and up to 512 GB of memory. With the optional external I/O expansion
unit, you can increase the number of supported I/O slots to a maximum of 112 PCIe and PCI-X slots.
• 16 Domains
• 1 Cabinet (700kg)
• 2 XSCFU (eXtended System Control Facility Unit) Service Processor
• 4 CMU (CPU Memory Boards) – 4 CPU sockets and 32 DIMM Slots per board
• 16 CPUs - SPARC64 VI dual-core and/or SPARC64 VII quad-core (2.28, 2.4 GHz or 2.52),
therefore up to 64 cores and 128 threads
• 512gb RAM (128 DIMM Slots using 4GB DIMMs)
• 4 IOU (8 PCIe slots & 4 Internal HDD)
• 16 Internal HDD
• 32 Internal PCIe slots
• Up to 8 External I/O Expansion Units
• Up to 112 PCI or PCI-X slots with the optional External I/O Expansion Unit
• 12 fan units (N+1 redundant), 4 x 172 mm (FAN_A) fans, 8 x 60 mm (FAN_B) fans
• Option Dual Power Feed (DPF) Single Phase Power Expansion Unit (7RU)
• Can be housed in the Base Cabinet
• Optional power expansion cabinet required for 3-phase power distribution (350kg)
Note that the OPL product line (M4000 - M9000) only supports Solaris 10_u3 11/06 and above. In order to
support COD and DR you will need Solaris 10_u4 8/07
The Sun SPARC Enterprise M9000 server features the option to mix and match up to 64 quad-core
SPARC64 VII processors or dual-core SPARC64 VI processors, with up to 256 cores, and 512 threads
per system
The Sun SPARC Enterprise M9000-32/64 server can support up to 24 dynamic domains regardless
whether it is connected to the expansion cabinet. For pure flexibility, we extended the domain partitioning
granularity down to one processor chip, four DIMMs and two I/O slots. An all-new, faster backplane
interconnect gives you 368 GB/second speeds and acts as a common interconnect between all chips,
maximizing address and data bandwidth on one interconnect.
The I/O system interface features hot-swappable I/O Units (IOUs), each of which provides eight PCIe
slots and four 2.5-inch SAS disk drive bays. Fully configured, these servers support up to 32 or 64
internal SAS boot disks, 64 or 128 internal PCIe slots, and up to 1TB or 2TB of memory (with 4GB DIMM).
With the optional external I/O expansion unit, you can increase the number of supported I/O slots to a
maximum of 224 slots (M9000-32) or 288 slots (M9000-64)
The M9000 server with Expansion Cabinet includes a total of 64 interconnect cables between the two
frames. Of the total number of cables, 62 of the cables are data and clock cables, one is an XSCFU_B
board cable, and one is an XSCFU_C board cable. Although the system cabinets are mostly the same,
there are small differences between the two sides. Therefore, one side is referred to as the main cabinet
and the other side is called the expansion cabinet.
• 1 Cabinet (940kg)
• 24 Domains
• 2 XSCFU_B (eXtended System Control Facility Unit) Service Processor
• 8 CMU (CPU Memory Boards) – 4 CPU sockets and 32 DIMM Slots per board
• 32 CPUs - SPARC64 VI dual-core and/or SPARC64 VII quad-core (2.28, 2.4 GHz or 2.52),
therefore up to 128 cores and 256 threads
• 1TB RAM (256 DIMM Slots using 4GB DIMMs)
• 8 IOU (8 PCIe slots & 4 Internal HDD)
• 32 Internal HDD
• 64 Internal PCIe slots
• Up to 16 External I/O Expansion Units
• Up to 224 PCI or PCI-X slots with the optional External I/O Expansion Unit
• 16 fan units (N+1 redundant), 16 x 172 mm fans
• 15 PSU (N+1 redundant)
• Optional power expansion cabinet required for 3-phase power distribution (350kg)
• 2 Cabinets (1880kg)
• 24 Domains
• 2 XSCFU_B (eXtended System Control Facility Unit) Service Processor
• Redundant XSCFU_B boards in the main cabinet. The XSCFU_C boards expansion
cabinet do not have any intelligence, they are used as a repeater to keep both
cabinets synchronized.
• XBU (System Crossbar) uses 16 individual XBU boards
• 8 XBU boards are plugged into the main cabinet’s centerplane
• 8 XBU boards are plugged into the expansion cabinet’s centerplane.
• The 2 sets of 8 expander boards are interconnected with 64 data and clock cables.
• 16 CMU (CPU Memory Boards) – 4 CPU sockets and 32 DIMM Slots per board
• 64 CPUs - SPARC64 VI dual-core and/or SPARC64 VII quad-core (2.28, 2.4 GHz or 2.52),
therefore up to 256 cores and 512 threads
• 2TB RAM (512 DIMM Slots using 4GB DIMMs)
• 16 IOU (8 PCIe slots & 4 Internal HDD)
• 64 Internal HDD
• 128 Internal PCIe slots
• Up to 16 External I/O Expansion Units
• Up to 288 PCI or PCI-X slots with the optional External I/O Expansion Unit
• 32 fan units (N+1 redundant), 32 x 172 mm fans
• 30 PSU (N+1 redundant)
• Optional power expansion cabinet required for 3-phase power distribution (350kg)
4. Components
The Mx000 Server Range was introduced in the earlier Overview Section, but before we start looking at the
components, here follows a recap of the characteristics:
On the M3000/4000/5000 )Mid-range) Servers, System Boards are not hot-pluggable and therefore the platform
needs to be powered off before a component can be replaced. On the M8000/M9000 Servers, the system
Boards can be replaced while the Platform is powered on. PCI cards for certain PCIe (PCI Express) and PCI-X
(PCI eXtended) controllers are hot-pluggable.
The Fan Units and Power Supplies on all M-Series Servers are Hot Pluggable meaning that these components
can be removed and inserted (replaced) without the need to power the Platform off.
Dynamic Reconfiguration (DR) is the process of replacing components while the O/S is running . PCI cards can
be replaced using the Solaris cfgadm command, If System Boards are hot-pluggable, they must be replaced
using the addfru, deletefru or eplacefru command from the XSCF.
At time of writing, the project on which I was working did not use Hot Swap and therefore this User Guide does
not cover the removal of components using DR.
4.3. PSU
Power supply units are hot replacement components on all M-Series Servers. Server PSU configurations are as
follows:
Power is supplied by redundant hot-swap power supplies, helping to support continued server operation even if a
power supply fails. Since the power units are hot-swappable, removal and replacement can occur while the
system continues to operate.
Note:
XSCF> showdualpowerfeed
Dual power feed is disabled.
4.4. FANU
Sun SPARC Enterprise M4000, M5000, M8000, and M9000 servers comprise redundant cooling configurations.
If a single fans fails, the XSCF detects the failure and switches the remaining fans to high-speed operation to
compensate for the reduced airflow. Each Server can operate normally under these conditions, allowing ample
time to service the failed unit. All fan units are hot pluggable, therefore replacement of fans units can occur
without interrupting application processing.
Note: Fan Trays must be replaced by running replacefru from the XSCF
More detail can be found in the Internal Sun Document Sun SPARC Enterprise[R] Mx000 server fan/fantray
redundancy and thermal environments
The 172-mm fan module is the primary cooling system. The 172-mm fan module is in both
midrange servers.
The 60-mm fan module is for additional cooling in the SPARC Enterprise M4000 server.
• M4000 (FF1)
• M5000 (FF2)
• M8000 (DC1)
• M9000/M9000+ (DC2/DC3)
4.5. MBU
On a Form Factor Server, The CPUM's and MEMB's plug into slots on the MBU (Mother Board or Main Board).
The MBU provides an Interconnect between CPU, Memory and I/O through the System Controller (SC) ASIC's
housed on the MBU. Both the M4000 and the M5000 have a single MBU but they both take a different number
of components:
• The M4000 MBU accepts up to 2 CPUM's, which allows up to 4 CPUs and 4 MEMB's which allow up to
32 DIMM's.
• The M5000 MBU accepts up to 4 CPUM's, which allows upto 8 CPUs and 8 MEMB's which allow up to
64 DIMM's.
Note that the High End Servers do not have any MBU's, instead they have several Backplane's.
4.6. Backplane
The DataCenter Servers (M8000/M9000) do not have a MBU, instead they have a Backplane.
4.6.1. BP_A
The M8000 has an Active Backplane with integrated Cross Bar (XB) providing the Clock (CLK)
4.6.2. BP_B
The M9000 has a Passive Backplane and additional Clock Boards (CLKU) which provide the Clock for
the ASIC's
4.7. XBU
The XBU (Cross Bar Unit) is only found on the M9000 and they provide connectivity to the M9000 Expansion
Cabinet which forms the M9000+. There are 8 XBU's in each cabinet which are connected by 64 Cables.
4.8. CLKU
CLKU (Clock Board Units) are only found in the M9000 and provide the Clock for the ASIC's. On an M8000 the
Clock (CLK) is integrated on the Active Backplane
To provide increased reliability, most of the internal parts of Sun SPARC Enterprise M9000 server clock chip are
redundant. In addition, there are two sources of clock signal and a dual signal line is implemented between the
clock chip and the system boards. By implementing redundant bus routes, the Sun SPARC Enterprise M9000
server can automatically restart in the event one route fails. The surviving route is used to continue operation.
The Form Factor Servers house CPU's and RAM on separate boards. CPU's are housed on the CPU Module
(CPUM) containing two CPU sockets which can host either SPARC64 VI or VII processor chips. On the
M4000 and M5000, there is no RAM on the CPUM, this is located on the Memory Board (MEMB) which can
contain up to 32 Memory DIMM's.
The Data Center Servers have CPU & RAM installed on the Same CMU (CPU Memory Board Unit)
Note: CPU & Memory information can be seen from the XSCFU by running the showhardconf command
Server No. of No. of No. of Max No. Max. No. of Max No. Max
CPUM MEMB CMU of Cores of RAM (GB)
Boards Boards Boards Procs Threads
M3000 N/A N/A N/A 1 4 8 32
M4000 2 4 N/A 4 16 32 128
M5000 4 8 N/A 8 32 64 256
M8000 N/A N/A 4 16 64 128 512
M9000-32 N/A N/A 8 32 128 256 1024
M9000-64 N/A N/A 16 64 256 512 2048
Notes:
• Maximum RAM capacity for Form Factor Servers is calculated based on 4-Gbyte DIMM
• Maximum RAM capacity for Data Center Servers is calculated based on 4-Gbyte DIMM
• Maximum CPU statistics for all Servers is based on the quad-core SPARC64 VII CPU (Jupiter)
• Memory patrol — Memory patrol periodically scans memory for errors. This proactive function
prevents the use of faulty areas of memory before they can cause system or application
errors, improving system reliability.
The memory subsystem on the Sun SPARC® Enterprise Mx000 (OPL) Servers is a complex
architecture. It is physically composed of 3 main components :
4.9.1.1. SC
The System Controller (SC) is a multi-function ASIC which directs traffic between local CPUs,
memory, I/O subsystems, and interconnect paths and provides cache coherency control.
There are 4 SC units on the CMU of the M8000/M9000 systems and 2 SC units on the MBU of
the M4000/M5000 systems. The physical addressing of all memory on each SB or CMU is
divided equally across the SCs on that board. The real memory is controlled by the MAC
(Memory Access Controller) under the SC.
4.9.1.2. MAC
The Memory Access Controller (MAC) is the chip which controls the memory access according to
the direction from the System Controller. The MAC is mounted on the CMU of the M8000/M9000
systems and on the MEMB of the M4000/M5000 systems.
4.9.1.3. DIMM's
The Servers support the following DIMM (Dual Inline Memory Module) sizes:
• 1GB DDR2 1-Rank DIMM (Form Factor & Data Center Servers)
• 2GB DDR2 1-Rank or 2-Rank DIMM (Form Factor & Data Center Servers)
• 4GB DDR2 2-Rank DIMM (Form Factor & Data Center Servers)
Note:
XSCF> showhardconf
SPARC Enterprise M5000;
+ Serial:BCF0904031; Operator_Panel_Switch:Locked;
+ Power_Supply_System:Single; SCF-ID:XSCF#0;
+ System_Power:On; System_Phase:Cabinet Power On;
Domain#0 Domain_Status:OpenBoot Execution Completed;
Additional information on Memory Configuration can be found by clicking in the following links:
4.9.2. CPUM
Each CPUM can contain 2 CPU sockets, he CPUM's are cold-FRU removal components:
• M4000 (FF1)
• M5000 (FF2)
4.9.3. MEMB
Each memory board provides a MAC (Memory Address Controller) and eight DIMM slots. The MEMBs
are cold-FRU replacement components. The entire system must be powered down and the power cords
disconnected to replace the memory boards.
• M4000 (FF1)
• M5000 (FF2)
• The size of the DIMMs in Group B must be smaller than or equal the size of the DIMMs
in Group A.
• The number of DIMMs in Group B must be either 0 or the same as the number of
DIMMs in Group A
• In UNI-XSB mode, MEMBs must be installed in powers of 2 (1,2 or 4). 3 MEMBs is not
possible, this is due to the single interleaving scheme used on an XSB.
• The minimum requirement for UNI-XSB mode is CPUM#0 / MEMB#0 (and CPUM#2 /
MEMB#4)
• A CPU on a CPUM without its associated MEMB and vice-versa cannot be used in
QUAD-XSB mode.
• Filling group B does not increase the interleaving factor, thus does not increase
memory performance. Only the memory capacity is increased.
4.9.4. CMU
The High-End Data Center Servers have a CPU/Memory Board Unit (CMU) which is equipped with
multiple CPU modules and memory. There are currently two types of CMUs available:
• CMU Type-A – A fully populated CMU board; contains four CPUs (8 or 16 cores)
• CMU Type-B – A partially populated CMU board; contains two CPUs (4 or 8 cores)
• M8000 (DC1):
• M9000 (DC2):
• M9000+(DC3):
• The size of the DIMMs in Group B must be smaller than or equal the size of the DIMMs
in Group A.
• The number of DIMMs in Group B must be either 0 or the same as the number of
DIMMs in Group A
• Memory must either be fully populated (32 DIMMs) or half populated (16 DIMMs).
• Mixing DIMM sizes within the same group is not supported. If populating DIMMs of
different sizes, make sure to install the larger value DIMMs in Group-A slots.
Memory Interleaving can be seen by running the prtdiag command as follows, in this example, 4-way
interleaving:
# prtdiag
System Configuration: Sun Microsystems sun4u Sun SPARC Enterprise M5000 Server
System clock frequency: 1012 MHz
Memory size: 16384 Megabytes
XSCF> setupfru -m y -x 1 sb 0
In Sun SPARC Enterprise M4000 and M5000 servers, memory is mirrored within the same memory
module, using the common memory address controller (MAC) Application Specific Integrated Circuit
(ASIC).
On Sun SPARC Enterprise M8000 and M9000 servers, memory is mirrored across adjacent MAC ASICs
to increase reliability. However, in a Quad-XSB configuration, paired DIMMs are split across different Sun
SPARC Enterprise M8000 and M9000 Quad-XSBs. As such, you can not configure Memory Mirroring on
the High End M8000/M9000 Server XSB in Quad-mode.
• SPARC64 VI Compatible Mode – All processors in the domain – which can be SPARC64 VI
processors, SPARC64 VII processors, or any combination of them – behave like and are treated
by the OS as SPARC64 VI processors. The new capabilities of SPARC64 VII processors are not
available in this mode.
• SPARC64 VII Enhanced Mode – All boards in the domain must contain only SPARC64 VII
processors. In this mode, the server utilizes the new features of these processors.
The operation mode can be changed using the setdomainmode(8) command, which is covered later in
this User Guide. To check the CPU operational mode, execute the prtdiag(1M) command (below) on the
Solaris OS. If the domain is in SPARC64 VII Enhanced Mode, the output will display SPARC64-VII on the
System Processor Mode line. If the domain is in SPARC64 VI Compatible Mode, nothing is displayed on
that line.
# prtdiag
System Configuration: Sun Microsystems sun4u Sun SPARC Enterprise M5000 Server
System clock frequency: 1012 MHz
Memory size: 32768 Megabytes
SPARC64-VII mode
4.10. I/O
The MX000 servers provide I/O functionality using a component called the IOU (I/O Unit). Each IOU has one I/O
controller, which manages four PCI buses. The IOU also contains cassettes that support PCI cards.
In order to facilitate hot-plug of PCIe (PCI Express) and PCI-X (PCI eXtended) adapter cards, Sun SPARC
Enterprise M4000, M5000, M8000, and M9000 servers utilize PCI cassettes. PCI cards which support PCI Hot
Plug can be mounted by administrators into a PCI cassette and inserted into an internal PCI slot or External I/O
Expansion Unit of a running server.
Server No. of IOU No. Of Max No. of Max No. Max No.
(I/O Units) Internal External of I/O of
I/O Slots I/O Units Slots Domains
M3000 N/A 4 Not Supported 4 1
M4000 1 5 2 25 2
M5000 2 10 4 50 4
M8000 4 32 8 112 16
M9000-32 8 64 16 224 24
M9000-64 16 128 16 288 24
• PCI-X (PCI eXtended) was developed by HP and IBM to improve PCI Bandwidth and is
backward compatible with existing PCI cards. The PCI-X implementation on OPL Servers allows
a clock speed of 133MHz, providing 1Gbyte/sec bandwidth.
• PCIe is a serial implementation of PCI & PCI-X which provide 8 lanes on the OPL Servers. Each
lane allows 250Mbytes/sec uplink and 250 Mbytes/sec downlink (therefore 500Mbytes/sec) and
therefore a total of 4Mbytes/sec Bandwidth with all 8 lanes. In fact, PCI Express was developed
to accommodate high-speed interconnects such as Fibre Channel, InfiniBand, and Gigabit
Ethernet.
PCI Bandwidth
4.10.2. IOU
I/O is added to the SPARC Enterprise M-series servers through assemblies called I/O Units (IOUs) that
contain a number of slots for PCI cards.
The Form Factor Servers share common I/O Technology with 1 or 2, 5 Slot IOU. The M4000 has one
IOU that provides four PCIe (PCI Express) slots and one PCI-X (PCI eXtended) slot. The IOU also has
several on-board devices that don’t take up any slots: two 1 gigabit Ethernet ports, a controller for two
Serial Attached SCSI (SAS) 2.5-inch disk drives, and a serial ATA (SATA) controller for the DVD and
optional DAT drive. The drives are installed in the system chassis, but the IOU is needed to provide
connectivity to the rest of the system.
The Sun SPARC Enterprise M5000 server can house up to two of these IOUs. The second IOU
contains the same 5 slots (4 PCIe & 1 PCI-X) and is required for connectivity to the third and fourth
built-in disk drives. There is a fixed relationship between the IOU and specific CPU/memory modules.
The first IOU is associated with CPUM 0 and 1, and MEMB 0 through MEMB 3. The second IOU is
associated with CPUM 2 and 3, and MEMB 4 through MEMB 7.
The M8000 and M9000 Data Center Server IOUs have eight PCIe slots (no PCI-X) and can hold up to
four 2.5-inch SAS disks in a 2 x 2 configuration. Unlike the IOUs in the Sun SPARC Enterprise M4000
and M5000 servers, there are no on-board controllers. The IOU Adapter card (IOUA) provides two 1
gigabit Ethernet ports and a SAS controller for up to two drives. IOUA card can be placed in any of the
even numbered PCIe slots. However, an IOUA is required in slot 0 in order to utilize disks 0 and 1, and
slot 4 for disks 2 and 3. The system chassis contains a DVD drive and optional DAT drive. An IOU with
an IOUA is required to access these peripherals.
• The single I/O boat configuration provides six slots for I/O cards which require 1 internal I/O
slot
• The optional two I/O boat configuration (dual boat) provides twelve slots. Each of the I/O
boats requires its own link kit, so the host server must have two I/O slots available for this
purpose.
If an I/O Unit is attached, you would see the output similar to that below from the following commands on
the XSCF:
XSCF> showhardconf
<... output deleted ...>
PCI#2 Status:Normal; Name_Property:; Card_Type:DownLink;
+ Ver:25h; Serial:XF08MP; Type:Optic;
+ Connection:IOX@X2K6/IOB0;
+ FRU-Part-Number:CF00501-7040 08/5017040-08;
IOX@X2K6 Status:Normal; Serial:XCX2K6;
+ FRU-Part-Number:CF00501-6937 07/5016937-07;
IOB0 Status:Normal; Serial:XE02VG; Type:PCI-Express;
+ FRU-Part-Number:CF00501-6939 06/5016939-06;
LINK Status:Normal; Ver:25h; Serial:XF08RK; Type:Optic;
+ FRU-Part-Number:CF00501-7040 08/5017040-08;
SLOT1 Name_Property:SUNW,emlxs;
SLOT2 Name_Property:SUNW,emlxs;
SLOT3 Name_Property:network;
SLOT4 Name_Property:network;
PS0 Status:Nornal; Serial:DD3227;
+ FRU-Part-Number:CF00300-1701 06/3001701-06;
PS1 Status:Normal; Serial:DD3254;
+ FRU-Part-Number:CF00300-1701 06/3001701-06;
<... output deleted ...>
In the Sun Fire Servers, this unit was commonly called the System Controller, but on an OPL Server, the System
Controller is a completely different component to the service processor (or XSCFU). On the OPL Servers, a
System Controller is a multi-function ASIC which directs traffic between local CPUs, memory, I/O subsystems,
and interconnect paths.
The XSCF is the firmware running on the Service Processor in the server. The board with the installed XSCF
firmware is called the XSCFU (also referred to as the "XSCF Unit") or Service Processor. So, what will we call
it? Well, we shall refer to it as the Service Processor or XSCFU in this document.
The XSCFU can configure the system to not use the broken part of the chip, and while the system remains
available, allowing you to schedule maintenance at a convenient time without incurring costly system down time.
The XSCFU enables standard control and monitoring functionality through the network.
The M3000, M4000 & M5000 use a single XSCFU, but the M8000 & M9000 use a redundant configuration of
XSCFU's, thereby providing high system reliability. The XSCFU that controls the server is called the Active
XSCFU, while the other XSCFU acts as a backup and is called the Standby XSCFU. The Active XSCFU and the
Standby XSCFU monitor each other, and if an error is detected, they determine when a failover switching to
Active or Standby should be performed.
The eXtended System Control Facility Unit (XSCFU) is a Service Processor that continuously monitors
the system hardware configuration and health, domain configuration and status, error monitor, and
notification on the Mx000 OPL Servers.
There is a firmware upgrade procedure to upgrade to a later (or earlier) version. The XCP F/W contains
updated functionality and bug fixes. The current latest XCP version (March 2009) is version 1081
The M3000 XSCFU is not removable, but looks and functions like the Form Factor XSCFU below.
4.11.4.2. M8000/M9000
The High End Datacenter Servers (M8000/M9000) have a pair of removable service processor
(XSCFU), pictured below. The Serial Port is marked No. 6, and the the NIC's are marked No. 3 &
4. Note that NIC 0 (xscf#0-lan#0) is the left hand port as you look at it.
Note – The MX000 servers are shipped without any of the networks configured.
The XSCF External Administration Network makes use of several interfaces to handle the
communication for the service processors. Each service processor has two network ports for
external communication to
the customer network. The high-end service processors provide floating IP addresses in case of
service processor failover.
Note
• The XSCF0 interfaces apply to both the mid-range and high-end servers. The XSCF1 and
lan#0 & lan#1 interfaces apply only to the high-end servers.
• Each service processor should be configured so that IP addresses for the network
interfaces are on two different subnets. The IP address associated with xscf#0-lan#0
should be on a different subnet than xscf#0-lan#1.
As you will see in Administration Section of this User Guide, DSCP is configured from the XSCF
using either the setupplatform or setdscp commands and from the domain by ensuring that the
dscp SMF Service is running. More details can be found in the Troubleshooting Section.
Note
• The DSCP network should only be configured when there are no domains running. If a
change is made to the DSCP network while a domain is active, the domain must be
rebooted before the service processor can communicate with it.
The XSCF ISN allows the active service processor to exchange pertinent system management
information with the standby service processor so that if a failure occurs, the standby service
processor can take over as the new active service processor without any interruption of domain
service. Network Interfaces The XSCF Internal Service Network consists of one ISN interface for
each service processor, and provides for a private internal communication.
• On the M4000 & M5000, the panel is located on the upper right front of the server.
• On the M8000 & M9000, the panel is located on the upper centre front of the Server
The Security Mode switch can be put into one of 2 settings using the OPNL key. The recommended setting for
all Mx000 Servers is Locked. Both the XSCP showhardconf command and the Solaris prtdiag command display
the keyswitch setting.
• Locked - This mode sets the server for normal operation, enabling error notifications.
• Service - This mode disables error notification as components are removed.
The 3 LED status indicators on the operator panel provide the following information:
The picture below is of a Data Center Operator Panel, which is horizontal (the Form Factor Operator Panels are
Vertical):
When the server is running, the Power and XSCF STANDBY LEDs (green) should be lit and the CHECK LED
(amber) should not be lit. If the CHECK LED is lit, search the system logs to determine what is wrong.
The Sun Rack 1000 hardware is designed to enclose Sun servers, storage systems, and qualified third-party
products. The Sun Rack 1000 hardware supports as many as three M5000 and six M4000 servers. It has a
redundant power configuration with two 32A IEC 309 five pin for 230/400 VAC three phase, pre-installed with the
rack
Sun Rack 1000 Hardware With Sun SPARC Enterprise M4000 and M5000 Servers
5. Concepts
5.1. Partitioning
OPL servers have a unique partitioning feature that allows the Administrator to configure a Physical System
Board (PSB) into 1 or 4 eXtended System Boards (XSB) using the setupfru command on the XSCF. These can
then be allocated to domains by running the setdcl & addboard commands
Unlike the Sun Fire range of Servers (F3800, F4800, F6800, F12K/F15K or E20K/E25K), an OPL Server needs
to be partitioned by the Administrator before any domains can be created. Basically, this means that the Server
must be configured so that the required number of domains can be created with the required resources (CPU,
RAM & I/O). If this is not done correctly, we may not be able to create the required number of domains, or the
domains created could have less resources than required. If it was configured incorrectly, you would then need
to remove any domains you have created, then remove the incorrect partitions and try again!
So in summary, Partitioning, or the creation of XSB's, is performed by the Administrator by logging into the
XSCF as a user with appropriate privileges and by then running the setupfru command. At the end of this
section, we will hopefully understand the configuration rules and options for the OPL Server range:
A Physical System Board (PSB) consists of up to 4 CPU's, 32 DIMM's and optional I/O. For example, on an fully
populated M4000, the 2 CPUM (CPU Modules) provide us with 4 CPU's and the 4 MEMB (Memory Boards)
provide us with 32 DIMMS, as shown in the diagram below (no I/O in this example):
Therefore, on an M4000, it is true to say that all these resources are automatically grouped into a single PSB:
If we only had 1 x CPUM or 2 x MEMB, all resources would still be allocated to 1 PSB, however, it would not be a
fully populated PSB.
The next step is to partition the Server. This is done by dividing a PSB into either 1 or 4 XSB's (eXtended
System Boards) through the Partitioning feature of these Servers.
A PSB (and any I/O slots in the associated IOU, if present) that is logically configured into one XSB is called a
Uni-XSB, whereas a PSB that is logically divided into four boards is called a Quad-XSB. Each resulting part of
the divided PSB (either 1 part or all 4 parts) is called an (XSB, regardless of it's size. These XSBs can be
combined freely to create domains, subject to certain restrictions.
A domain requires, at the very least, a CPU, some memory and some I/O (boot disk & network) to be of any use.
You create a domain by assigning XSB's to it, but first you must create the XSB's. You will also see that I/O
resources on an IOU are tied to a specific PSB, and that CPU, Memory, & I/O resource are automatically
assigned to a XSB.
The minimal hardware required for a functioning XSB is to have at least 1 CPU and at least 4 DIMM's, the
maximum hardware which will be allocated to an XSB is 4 CPU's and 32 DIMM's. Therefore, a domain can
consists of 1 or more XSB,s but at least 1 of them must have CPU, memory and I/O. You will see, as we move
on, that I/O is not always allocated to an XSB, but that is OK. Provided the XSB has a valid configuration of CPU
and Memory, it can still be allocated to a domain.
• If there is CPU and no Memory, then the resulting XSB is not usable
• If there is Memory and no CPU, then the resulting XSB is not usable
If an XSB does not meet minimum configuration requirments, it will appear as status “unmount” in the
Test column when you run showboards, as follows:
XSCF> setupfru -x 4 sb 0
Operation has completed. However, a configuration error was detected.
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Quad no
So, in summary, a PSB can be divided into uni-XSB (default) or quad-XSB from the XSCF using the setupfru
command. The resulting objects are referred to as XSB's
5.3.2. Uni-XSB
When a PSB (and any I/O slots in the associated IOU, if present) is configured as a Uni-XSB, this results
in a single XSB. All resources associated with that PSB and IOU are placed in a single XSB and can only
be allocated to one domain, not shared between domains.
For example, the M4000 only supports one domain when the PSB is in the Uni-XSB configuration. In this
case, all components of the single PSB (PSB0) are allocated to a single XSB (XSB 00-0), which in turn is
allocated to a single domain (DID 0).
The above example would be setup from the XSCF by typing the following command. The output will be
discussed in a later section.
XSCF> setupfru -x 1 sb 0
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Uni no
5.3.3. Quad-XSB
When a PSB (and any I/O slots in the associated IOU, if present) is configured as a Quad-XSB, this
results in 4 XSB's. As, the minimal hardware required for a functioning XSB is to have at least 1 CPU and
at least 4 DIMM's, so each of the 4 resulting XSB's must have this minimum configuration to be usable.
In this case, a PSB is broken down into smaller parts which can then be allocated to multiple domains.
Provided they meet the minimum usable requirements, each of the 4 resulting XSB's can be assigned to
a domain. All resources associated with resulting XSB and can only be allocated to one domain, not
shared between domains.
For example, the M4000 only supports one domain when the PSB is in the Uni-XSB configuration, but it
can support 2 Domains when in Quad-XSB configuration. In this case, all components of the single PSB
(PSB0) are divided between the resulting 4 XSB's. However, only 2 domains can be created because
only 2 of the resulting XSB's will have I/O slots allocated.
Probably the most important thing to know when configuring Quad-XSB is that for a resulting XSB to be
usable, a CPU must be present and there must be at least 4 DIMMS in Group A of the Memory Board at
that location. If there is no CPU present, the Memory is not usable, even if it exists. If there is no
Memory, the CPU at that location is not usable, even if it exists.
The above example would be setup from the XSCF by typing the following command. The output will be
discussed in a later section.
XSCF> setupfru -x 4 sb 0
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Quad no
If the capability to switch between UNI-XSB and QUAD-XSB mode is desired, then both the rules for UNI-
XSB and those for QUAD_XSB must be followed. The supported M8000 / M9000 configurations enforce
this.
UNI-XSB mode allows higher memory performance because it can use higher interleave factors. If all
resources of a PSB are needed in one domain, then it is advisable to configure the PSB in UNI-XSB
mode.
Memory mirroring is an effective way to protect against memory failures. Configuring an XSB for memory
mirroring halves the available memory capacity. It also halves the interleave factor. Memory mirroring
should therefore best be used in UNI-XSB mode.
The DCL provides a mechanism for the platform administrator to control which XSB resources can be used in a
particular domain. It is possible for an XSB to appear in multiple domain's DCL, allowing the XSB to be moved
between domains, for example, using Dynamic Configuration (DR). An XSB is can be active in only one domain
at at a time.
However, registering an XSB with the DCL also performs another very important function, it allows us to assign a
Logical System Board (LSB) against the XSB. This LSB number (0 – 15) is used in any component device
name, either through the OBP or Solaris.
A single XSB can belong to the DCL of multiple domains, but can not be in more than one domain at a time. A
good example of this would be if you wanted to frequently move an XSB between 2 domains. The XSB would
exist in the DCL for both Domains and could be moved as required using either static configuration, or perhaps
Dynamic Reconfiguration, if supported.
This option is only relevant if you intend using Dynamic Reconfiguration (DR) to remove the XSB.
At the time of writing (Feb 2009) and for the foreseeable future, Annuity does not support the use
of DR to remove boards.
For diagnosis and management of a system board, memory must be mounted on the system
board even if the omit-memory option is enabled. The value of this option is “true” (omit memory)
or “false” (do not omit memory). The default value is “false”.
We have previously seen a fully populated M4000 used as an example to demonstrate allocating all resources
from PSB0 and IOU0 to a single Uni-XSB (called 00-0). Building on that example, we register the XSB with LSB
0 with the DCL for Domain 0, and then configure it not to use memory for LSB#00 of domain 0:
Each domain has its own set of LSB assignments. Unlike an E25K for example, where every device in the frame
has a unique device name dependent on the System or I/O Board on which it reside, the device names on the
Form Factor and Data Center Servers is dependent on the LSB No. which is assigned to the XSB.
In short, the specific purpose of the LSB layer is “tell” the OS how you want it to “see” the resource, though the
actual “resource” assigned to the domain is done at the XSB layer. In summary, there are two steps:
• Assign the XSB resource to an LSB to tell the OS how you want the system to display/use/see the
resource if/when it is configured into the domain (by using the setdcl command)
• Assign the XSB resource to the domain to tell the domain to use the resource (by using the addboard
command)
The nice thing about the "abstraction" that the LSB provides is that it allows us to logically move an XSB between
domains without having to physically move them. We do this by assigning the XSB to multiple DCL's using the
same LSB No. for each one.
The LSB information for a domain can be seen by running the showdcl command. The following output builds on
the example where we want to create a single domain on a fully populated M4000 with a single Uni-XSB:
• The LSB can be seen both from the XSCF (showdcl below)
XSCF> showdcl -v -d 0
DID LSB XSB Status No-Mem No-IO Float Cfg-policy
00 Running FRU
00 00-0 False False False
01 -
02 -
03 -
04 -
05 -
06 -
07 -
08 -
09 -
10 -
11 -
12 -
13 -
14 -
15 -
# prtdiag
System Configuration: Sun Microsystems sun4u Sun SPARC Enterprise M5000 Server
System clock frequency: 1012 MHz
Memory size: 65536 Megabytes
==================================== CPUs ====================================
SPARC64-VII mode
5.6. Domains
A domain is an independent system resource that runs its own instance of the Solaris OS. Operations in one
domain are not affected by operations in another domain. Domains can be used to perform different types of
processing activity. For example, one domain can be used to test new applications, while another domain can be
used for production purposes.
So, why is the M9000-64 limited to only 24 domains if It has 64 cpus? Actually, this has nothing to do with the
number of cpus. The system uses the JBUS to address domains. JBUS uses a 5-bit protocol. This means there
are a maximum of 25 or 32 possible domains. HOWEVER, the system reserves 8 domains to be used in running
post if an IOU is inserted into a domain. This leave 24 domains for the Administrator to use.
Domains are created by assigning XSB's which were previously configured by running the setupfru command
and registered with a DCL by running setdcl command. The following output builds on the example where we
want to create a single domain on a fully populated M4000 with a single Uni-XSB:
XSCF> showboards -a -v
XSB R DID(LSB) Assignment Pwr Conn Conf Test Fault COD
---- - -------- ----------- ---- ---- ---- ------- -------- ----
00-0 00(00) Assigned y y y Passed Normal n
Note that the documentation tells us that the XSB assigment state must be “Available” to be able to assigne an
XSB to a domain using addboard, but while writing this User Guide, I found that as long as the IDI(LSB) Couln
was “SP”, it did not matter if the assigment state was “Available” or “Unavailable”, it worked with either.
Once a domain has been created, it needs to be powered on so that POST can run and the OBP can be loaded,
after which we will want to start a console to access the OBP as follows:
XSCF> poweron -d 0
XSCF> console -d 0
Boards can also be removed from Domains. The FF & DC OPL Servers allow you to remove a board from a
Domain which is powered off (Static) and one which is powered on and running Solairs (Dynamic), both of which
are covered later in this document:
XSCF> poweron -d 0
6. Device Paths
The following section explains how to locate devices to a particular slot on an a physical CPU, Memory or I/O Board and
provides some working examples to check your understanding of I/O and CPU Device Paths for both FF and DC
Servers. By the end of the section you will have seen that the LSB number you choose for your domain is actually used
by the OBP & Solaris when identifying CPU's and I/O Devices, and that the LSB No. along with the DID allows you to
locate a device.
Some additional references to help with understanding Device Paths can be found as follows (these can also be found
in the Bookmarks Section of this User Guide):
Basically, look for the first number after the '/pci@' but before the comma, and the next number after the
comma (e.g /pci@0,600000). If the 2nd number is a 600000 or 700000 then we can tell the LSB#0 of the
board which this device belongs too:
• If the number is a 2-digit number (e.g 10 or 20), the first digit is the LSB# (e.g 1 or 2)
• If the number is a single digit number (e.g 0 or 2), then there is a “hidden” 0 and this is LSB#0
There are many ways to view I/O Device Names, a few of them are listed below:
• OBP
{0} ok show-nets
a) /pci@0,600000/pci@0/pci@8/pci@0/network@2,1
b) /pci@0,600000/pci@0/pci@8/pci@0/network@2
q) NO SELECTION
Enter Selection, q to quit: q
{0} ok show-devs
<... stuff deleted...>
/pci@0,600000/pci@0
/pci@0,600000/pci@0/pci@9
/pci@0,600000/pci@0/pci@8
/pci@0,600000/pci@0/pci@8/pci@0,1
/pci@0,600000/pci@0/pci@8/pci@0
/pci@0,600000/pci@0/pci@8/pci@0/network@2,1
/pci@0,600000/pci@0/pci@8/pci@0/network@2
/pci@0,600000/pci@0/pci@8/pci@0/scsi@1
/pci@0,600000/pci@0/pci@8/pci@0/scsi@1/disk
/pci@0,600000/pci@0/pci@8/pci@0/scsi@1/tape
<... stuff deleted...>
{0} ok show-disks
a) /pci@0,600000/pci@0/pci@8/pci@0/scsi@1/disk
q) NO SELECTION
Enter Selection, q to quit: q
{0} ok boot
Boot device: /pci@0,600000/pci@0/pci@8/pci@0/scsi@1/disk@0,0:a File and args:
SunOS Release 5.10 Version Generic_138888-02 64-bit
Copyright 1983-2008 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
# echo | format
Searching for disks...done
In the format example above, we use a fully loaded M5000 which comprises 2 Uni-XSB's (the showdcl &
prtdiag is a real view of the configuration). The device names can be read as follows:
• Disk 0 = /pci@0,600000
There is only one digit after the 'pci@' but before the comma, that number is a zero (0). The
numbers after the comma are 60000 so fall within the 600000 or 700000 range.
Therefore in this case, Disk 0 has a “hidden” zero (0) and this device is on LSB#0
• Disk 1 = /pci@10,600000
There are 2 digits after the 'pci@' but before the comma, and they are one-zero (10). The
numbers after the comma are 60000 so fall within the 600000 or 700000 range.
It is worth noting that on a FF Server, the CPU is not an FRU and so the whole CPUM (less memory) may need
to be replaced. On a DC Server, the CPU is an FRU. Provided we know the DID of the domain a CPU is in, we
will able to locate it.
Remember that a CPU can be dual-core or quad-core, and that each core has 2 strands. A CPU strand can be
identified by a decimal number (e.g Solaris or XSCF) or by a hexidecimal number (e.g OBP). It is the strand
which supports a S/W thread and it is the strand which has a unique identifier.
Here are some examples of the same 4 CPU strands (0-4), from 2 dual-core CPU's, each with 2 strands, on a
CPUM which is part of a uni-XSB (00-0) in a Domain (DID 0) of an M5000:
• OBP
{0} ok show-devs
<... stuff deleted...>
/cmp@400,0/core@1/cpu@1
/cmp@400,0/core@1/cpu@0
/cmp@400,0/core@0/cpu@1
/cmp@400,0/core@0/cpu@0
<... stuff deleted...>
• Solaris
# psrinfo
0 on-line since 02/27/2009 16:58:45
1 on-line since 02/27/2009 16:58:46
2 on-line since 02/27/2009 16:58:46
3 on-line since 02/27/2009 16:58:46
# prtdiag
System Configuration: Sun Microsystems sun4u Sun SPARC Enterprise M5000
System clock frequency: 1012 MHz
Memory size: 16384 Megabytes
# cfgadm -al
SB0 System_Brd connected configured ok
SB0::cpu0 cpu connected configured ok
SB0::cpu1 cpu connected configured ok
SB0::cpu2 cpu connected configured ok
SB0::cpu3 cpu connected configured ok
SB0::memory memory connected configured ok
<... stuff deleted ...>
• XSCF
XSCF> showdevices -d 0
CPU:
----
DID XSB id state speed ecache
00 00-0 0 on-line 2150 5
00 00-0 1 on-line 2150 5
00 00-0 2 on-line 2150 5
00 00-0 3 on-line 2150 5
Calculate this by taking the decimal CPU strand identifier (43), and divide by 32 discarding any
remainder, e.g:
CPU = 43
LSB = 43/2
LSB = 1.34375
LSB =1
Calculate this by taking the decimal CPU strand identifier, and divide by 32, but this time we are
only interested in the remainder, so we use modular arithmatic. Once we have the remainder, we
divide by 8 and discard any remainder:
CPU = 43
A = 43 MOD 32
A = 11
CPU = 11/8
CPU = 1.375
CPU =1
We now know the DID # (0), the LSB # (1) and the CPU # (1), so we now need to find the XSB
which will lead us to the PSB:
XSCF> showboards -d 0
XSB R DID(LSB) Assignment Pwr Conn Conf Test Fault COD
---- - -------- ----------- ---- ---- ---- ------- -------- ----
00-0 00(00) Assigned y y y Passed Normal n
01-0 00(01) Assigned y y y Passed Normal n
XSB =1
The 1st 2 digit's of the XSB (01-0), identify the PSB, so this is PSB 1
PSB =1
So, we can conclude that in this example, CPU # 43 (decimal) is actually CPU 1 on PSB 1. As
there are 2 CPU's per CPUM, this will be CPUM # 2.
XSCF> showdcl -a
DID LSB XSB Status
00 Running
00 00-0
---------------------------
01 Powered Off
00 01-0
XSCF> showdevices -d 0
CPU:
----
DID XSB id state speed ecache
00 00-0 0 on-line 2150 5
00 00-0 1 on-line 2150 5
00 00-0 2 on-line 2150 5
00 00-0 3 on-line 2150 5
00 00-0 8 on-line 2150 5
00 00-0 9 on-line 2150 5
00 00-0 10 on-line 2150 5
00 00-0 11 on-line 2150 5
00 00-0 16 on-line 2150 5
00 00-0 17 on-line 2150 5
00 00-0 18 on-line 2150 5
00 00-0 19 on-line 2150 5
00 00-0 24 on-line 2150 5
00 00-0 25 on-line 2150 5
00 00-0 26 on-line 2150 5
00 00-0 27 on-line 2150 5
Memory:
-------
board perm base domain target deleted remaining
DID XSB mem MB mem MB address mem MB XSB mem MB mem MB
00 00-0 16384 1327 0x000003c000000000 16384
IO Devices:
----------
DID XSB device resource usage
00 00-0 sd1 /dev/dsk/c0t0d0s0 component of concat "/dev/md/dsk/d11"
00 00-0 sd1 /dev/md/dsk/d11 submirror of "/dev/md/dsk/d10"
00 00-0 sd1 /dev/md/dsk/d10 mounted filesystem "/"
00 00-0 sd1 /dev/dsk/c0t0d0s1 component of concat "/dev/md/dsk/d21"
00 00-0 sd1 /dev/md/dsk/d21 submirror of "/dev/md/dsk/d20"
00 00-0 sd1 /dev/md/dsk/d20 swap area
00 00-0 sd1 /dev/dsk/c0t0d0s3 component of concat "/dev/md/dsk/d31"
00 00-0 sd1 /dev/md/dsk/d31 submirror of "/dev/md/dsk/d30"
00 00-0 sd1 /dev/dsk/c0t0d0s5 component of concat "/dev/md/dsk/d51"
00 00-0 sd1 /dev/md/dsk/d51 submirror of "/dev/md/dsk/d50"
00 00-0 sd1 /dev/md/dsk/d50 contains soft partition(s)
00 00-0 sd1 /dev/md/dsk/d80 soft partition based on "/dev/md/dsk/d50"
00 00-0 sd1 /dev/md/dsk/d80 mounted filesystem "/var/crash"
00 00-0 sd1 /dev/dsk/c0t0d0s7 contains metadb(s)
00 00-0 sd0 /dev/dsk/c0t1d0s7 contains metadb(s)
00 00-0 bge0 SUNW_network/bge0 bge0 hosts IP addresses: 10.131.12.23,
10.131.99.23
00 00-0 bge1 SUNW_network/bge1 bge1 hosts IP addresses: 10.131.99.24
# psrinfo
0 on-line since 02/27/2009 16:58:45
1 on-line since 02/27/2009 16:58:46
2 on-line since 02/27/2009 16:58:46
3 on-line since 02/27/2009 16:58:46
8 on-line since 02/27/2009 16:58:46
9 on-line since 02/27/2009 16:58:46
10 on-line since 02/27/2009 16:58:46
11 on-line since 02/27/2009 16:58:46
16 on-line since 02/27/2009 16:58:46
17 on-line since 02/27/2009 16:58:46
18 on-line since 02/27/2009 16:58:46
19 on-line since 02/27/2009 16:58:46
24 on-line since 02/27/2009 16:58:46
25 on-line since 02/27/2009 16:58:46
26 on-line since 02/27/2009 16:58:46
27 on-line since 02/27/2009 16:58:46
# prtdiag
System Configuration: Sun Microsystems sun4u Sun SPARC Enterprise M5000 Server
System clock frequency: 1012 MHz
Memory size: 16384 Megabytes
{0} ok
{0} ok cd /
{0} ok cd cmp@418,0
{0} ok ls
f0092d1c core@1
f0092694 core@0
{0} ok cd core@0
{0} ok ls
f0092bf0 cpu@1
f0092afc cpu@0
XSCF> showboards -a
XSB DID(LSB) Assignment Pwr Conn Conf Test Fault
---- -------- ----------- ---- ---- ---- ------- --------
01-0 07(01) Assigned y y y Passed Normal
01-1 01(05) Assigned y y y Passed Normal
01-2 00(02) Assigned y y y Passed Normal
01-3 03(02) Assigned y y y Passed Normal
03-0 03(12) Assigned y y y Passed Normal
03-1 07(03) Assigned y y y Passed Normal
03-2 01(04) Assigned y y y Passed Normal
03-3 07(09) Assigned y y y Passed Normal
XSCF> showdcl -a
DID LSB XSB Status
00 Running
02 01-2
---------------------------
01 Running
04 03-2
05 01-1
---------------------------
03 Running
02 01-3
12 03-0
---------------------------
07 Running
01 01-0
03 03-1
09 03-3
XSCF> showdevices -d 7
CPU:
----
DID XSB id state speed ecache
07 01-0 32 on-line 2280 5
07 01-0 33 on-line 2280 5
07 01-0 34 on-line 2280 5
07 01-0 35 on-line 2280 5
07 03-1 104 on-line 2280 5
07 03-1 105 on-line 2280 5
07 03-1 106 on-line 2280 5
07 03-1 107 on-line 2280 5
07 03-3 312 on-line 2280 5
07 03-3 313 on-line 2280 5
07 03-3 314 on-line 2280 5
07 03-3 315 on-line 2280 5
Memory:
-------
board perm base domain target deleted remaining
DID XSB mem MB mem MB address mem MB XSB mem MB mem MB
07 01-0 16384 2150 0x0000038000000000 49152
07 03-1 16384 0 0x0000030000000000 49152
07 03-3 16384 0 0x0000018000000000 49152
IO Devices:
----------
DID XSB device resource usage
07 01-0 sd1 /dev/dsk/c0t0d0s0 mounted filesystem "/"
07 01-0 sd1 /dev/dsk/c0t0d0s1 swap area
07 01-0 sd1 /dev/dsk/c0t0d0s1 dump device (swap)
07 01-0 bge0 SUNW_network/bge0 bge0 hosts IP addresses: 10.2.89.135
XSCF> console -d 7
# psrinfo
32 on-line since 03/10/2009 14:23:00
33 on-line since 03/10/2009 14:23:01
34 on-line since 03/10/2009 14:23:01
35 on-line since 03/10/2009 14:23:01
104 on-line since 03/10/2009 14:23:01
105 on-line since 03/10/2009 14:23:01
106 on-line since 03/10/2009 14:23:01
107 on-line since 03/10/2009 14:23:01
312 on-line since 03/10/2009 14:23:01
313 on-line since 03/10/2009 14:23:01
314 on-line since 03/10/2009 14:23:01
315 on-line since 03/10/2009 14:23:01
# init 0
{20} ok
7. Operation
7.1. Connectivity
7.1.1. Connecting to the XSCF
You can login to the XSCF over the Serial Port or over the network, the same as an E25K Controller.
• If you login over the Serial Port, you may get the Solaris Root Console Login instead of the
XSCF login , do a # . and this should get you down to the XSCF> prompt, then exit. The
restriction with a serial port login is that only one person can use the XSCF at a time.
Without a network connection, we would not be able to monitor the XSCFU
$ ssh user@tc
> conn mx000-sc
$ ssh mx000-xscf-userr@mx000-sc
$ spasswd xscf
The following example is of a connection to the Serial Port of the XSCF of an M5000. In this case, when
we connect to the XSCF, we get the XSCF login prompt. We login in, establish a console connection
with a domain, then logout again:
login: mx000-xscf-user
XSCF> showdomainstatus -a
DID Domain Status
00 -
01 -
02 -
03 -
XSCF> console -d 0
Connect to DomainID 0?[y|n] :y
XSCF> who
USER TTY IDLE FROM HOST
m5000-admin pts/1 00:00m Mar 12 14:39 195.121.217.14
XSCF> showconsolepath -a
User DID ro/rw escape Date
m5000-admin 0 rw # Thu Mar 12 14:47:00 2009
XSCF> exit
login:
Alternatively, you can type part of a command and then do a <TAB> <TAB> at the XSCF>
prompt and it will list all available commands:
XSCF> delete <TAB> <TAB>
deleteboard deletecodlicense deletefru deleteuser
XSCF> console -h
usage : console -d domain_id [[-q]{-y|-n}] [-f|-r] [-s escapeChar]
console -h
NAME
exit - exit the XSCF shell
SYNOPSIS
exit
DESCRIPTION
The exit(1) command exits and closes the XSCF shell.
Privileges
No privileges are required to run this command.
7.2. Display
7.2.1. Platform
The following commands display basic information about the platform configuration and status:
7.2.1.1. showhostname
Display XSCF hostnames
XSCF> showhostname -a
xscf#0:c1718-4-92-mgt-xscf0.ssclabs.net
7.2.1.2. shownetwork
The shownetwork command displays XSCF network configuration details:
XSCF> shownetwork -a
xscf#0-lan#0
Link encap:Ethernet HWaddr 00:21:28:25:DE:D8
inet addr:10.130.4.92 Bcast:10.130.255.255 Mask:255.255.0.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:2182410 errors:0 dropped:0 overruns:0 frame:0
TX packets:41151 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:136027156 (129.7 MiB) TX bytes:17908147 (17.0 MiB)
Base address:0xe000
xscf#0-lan#1
Link encap:Ethernet HWaddr 00:21:28:25:DE:D9
BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)
Base address:0xc000
7.2.1.3. showroute
The showroute command displays XSCF routing details:
XSCF> showroute -a
Destination Gateway Netmask Flags Interface
10.130.0.0 * 255.255.0.0 U xscf#0-lan#0
default 10.130.0.1 0.0.0.0 UG xscf#0-lan#0
7.2.1.4. showdscp
The showdscp command displays the IP addresses assigned to the Domain to Service Processor
Communications Protocol (DSCP)
XSCF> showdscp
DSCP Configuration:
Network: 192.168.224.0
Netmask: 255.255.255.0
Location Address
---------- ---------
XSCF 192.168.224.1
Domain #00 192.168.224.2
Domain #01 192.168.224.3
Domain #02 192.168.224.4
Domain #03 192.168.224.5
7.2.1.5. showsunmc
This command is used to show the setup information and current status of the Sun Management
Center agent.
XSCF> showsunmc
Agent Status: Disabled, stopped
Setup Status: Not set up
SunMC Server: NOT_SET
Security Seed: maplesyr
SNMPv1 Community String: public
Agent Port: 1161
Host Trap Port: 162
Host Event Port: 163
SNMP Agent Port: 161
Domain Agent Port: 1161
7.2.1.6. shownameserver
The shownameserver(8) command displays the registered DNS servers in the XSCF network.
XSCF> shownameserver
nameserver 10.131.2.2
7.2.1.7. showdate
The showdate(8) command displays the date and time of XSCF.
XSCF> showdate
Tue Feb 24 04:55:47 GMT 2009
7.2.1.8. showhardconf
The showhardconf command displays a variety of information about the H/W configuration of the
server and also the status of each Domain.
Pay special attention to the output and look for an asterisk (*) in the first column of the output as
this indicates a problem which requires attention. An example of this is included at the end of this
document where we follow through a PSU problem on an M4000.
XSCF> showhardconf
SPARC Enterprise M5000;
+ Serial:BCF0904031; Operator_Panel_Switch:Locked;
+ Power_Supply_System:Single; SCF-ID:XSCF#0;
+ System_Power:On; System_Phase:Cabinet Power On;
Domain#0 Domain_Status:OpenBoot Execution Completed;
PPMYYWWSSS
In the example above, serial number BCF0904031 indicates it is from week 04 of 2009
showhardconf -u
You can also view a formatted output with the -u option:
XSCF> showhardconf -u
SPARC Enterprise M5000; Memory_Size:64 GB;
+-----------------------------------+------------+
| FRU | Quantity |
+-----------------------------------+------------+
| MBU_B | 1 |
| CPUM | 4 |
| Freq:2.400 GHz; | ( 8) |
| MEMB | 4 |
| MEM | 32 |
| Type:2B; Size:2 GB; | ( 32) |
| DDC_A | 4 |
| DDC_B | 2 |
| IOU | 2 |
| DDC_A | 2 |
| DDCR | 2 |
| DDC_B | 2 |
| XSCFU | 1 |
| OPNL | 1 |
| PSU | 4 |
| FANBP_C | 1 |
| FAN_A | 4 |
+-----------------------------------+------------+
7.2.1.9. showenvironment
Display the intake air temperature and humidit, temperature sensor information, voltage sensor
information, and fan speed information about the system
7.2.2. Domain
The following commands display basic information about the domain configuration and status:
7.2.2.1. showboards
The showboards(8) command displays system board information including the domain ID of the
domain to which the target system board belongs and various kinds of system board status in list
format. It does not tell us the domain status (Running or OBP) or what mode the XSB's are in.
In this example, we have an M5000 with 2 XSB's. Domain 0 has 1 XSB assigned (XSB 00-0).
7.2.2.2. showfru
The showfru(8) command displays information about the PSB division type and memory mirroring
mode settings in list format. To change the PSB configuration, use the setupfru(8) command.
7.2.2.3. showdcl
The XSCF showdcl command displays domain information including the domain ID, configured
system board numbers, and domain status
In this example, we have an M5000 with 1 Domain (DID 0) which is at the OBP (after POST). The
domain comprises 1 XSB (XSB 00-0). There are no other DCL's configured.
7.2.2.4. showdomainstatus
The showdomainstatus(8) command lists the domains in the system and their status. This
command displays the same domain status information as the showdcl(8) command.
In this example, we have an M5000 with 2 Domains. Domain 0 is at the OBP (after POST) and
Domain 1 is Running Solaris.
XSCF> showdomainstatus -a
DID Domain Status
00 OpenBoot Execution Completed
01 Running
02 -
03 -
XSCF> showdomainstatus -a
DID Domain Status
00 Running
01 -
CPU:
----
DID XSB id state speed ecache
00 00-0 0 on-line 2400 5
00 00-0 1 on-line 2400 5
00 00-0 2 on-line 2400 5
00 00-0 3 on-line 2400 5
00 00-0 4 on-line 2400 5
00 00-0 5 on-line 2400 5
00 00-0 6 on-line 2400 5
00 00-0 7 on-line 2400 5
00 00-0 8 on-line 2400 5
00 00-0 9 on-line 2400 5
00 00-0 10 on-line 2400 5
00 00-0 11 on-line 2400 5
00 00-0 12 on-line 2400 5
00 00-0 13 on-line 2400 5
00 00-0 14 on-line 2400 5
00 00-0 15 on-line 2400 5
00 00-0 16 on-line 2400 5
00 00-0 17 on-line 2400 5
00 00-0 18 on-line 2400 5
00 00-0 19 on-line 2400 5
00 00-0 20 on-line 2400 5
00 00-0 21 on-line 2400 5
00 00-0 22 on-line 2400 5
00 00-0 23 on-line 2400 5
00 00-0 24 on-line 2400 5
00 00-0 25 on-line 2400 5
00 00-0 26 on-line 2400 5
00 00-0 27 on-line 2400 5
00 00-0 28 on-line 2400 5
00 00-0 29 on-line 2400 5
00 00-0 30 on-line 2400 5
00 00-0 31 on-line 2400 5
Memory:
-------
board perm base domain target deleted
remaining
DID XSB mem MB mem MB address mem MB XSB mem MB mem MB
00 00-0 32768 1624 0x000003c000000000 32768
IO Devices:
----------
DID XSB device resource usage
00 00-0 sd2 /dev/dsk/c0t0d0s0 mounted filesystem "/"
00 00-0 sd2 /dev/dsk/c0t0d0s1 swap area
00 00-0 sd2 /dev/dsk/c0t0d0s1 dump device (swap)
00 00-0 bge0 SUNW_network/bge0 bge0 hosts IP addresses:
10.128.3.92
7.3. Control
7.3.1. Platform
7.3.1.1. Reboot the XSCF
The XSCF can be rebooted with no impact on the running domains using the rebootxscf
command
XSCF> rebootxscf
The XSCF will be reset. Continue? [y|n] :y
7.3.2. Domain
7.3.2.1. Start the Domain
XSCF> showboards -av
XSB R DID(LSB) Assignment Pwr Conn Conf Test Fault COD
---- - -------- ----------- ---- ---- ---- ------- -------- ----
00-0 * 00(00) Assigned y n n Passed Normal n
XSCF> poweron -d 0
DomainIDs to power on:00
Continue? [y|n] :y
00 :Powering on
*Note*
This command only issues the instruction to power-on.
The result of the instruction can be checked by the "showlogs power".
XSCF> exit
7.3.2.4. showdomainmode
The showdomainmode(8) command displays the modes of operation that is set for the specified
domain:
XSCF> showdomainmode -d 0
Host-ID :8525ded4
Diagnostic Level :max
Secure Mode :on
Autoboot :on
CPU Mode :auto
7.3.2.5. setdomainmode
The setdomainmode(8) sets the modes of operation for the specified domain. The modes of
operation for the specified domain include the following types:
• Diagnostics Level
• Secure Mode
Whether to enable or disable the host watchdog and suppress break signal reception. The
default of the host watchdog is enable and suppress break signal reception is enable.
• Autoboot
Whether to enable or disable the auto boot function used at domain startup. The default is
enable.
• CPU Mode
Way of determining the CPU operational mode mounted on the domain. The CPU
operational mode can be automatically determined at domain startup, or manually set to
the compatible mode. The default is to let it automatically determined at domain startup.
7.3.2.6. setdomparam
The setdomparam(8) command rewrites OpenBoot PROM environment variables for a specified
domain. The following OpenBoot PROM environment variables can be specified.
• use-nvramrc?
Whether to execute the contents of the NVRAM at the boot or reboot of a domain.
• security-mode?
• set-defaults
Whether to restore OpenBoot PROM environment variables to the settings at the time of
shipment from the factory
Example 3, Initializes the OpenBoot PROM environment variables of DID 0 to factory settings
XSCF> setdomparam -d 0 set-defaults
DomainIDs of domains that will be affected:00
All OpenBoot PROM variable will be reset to original default values.
Continue? [y|n]:y
Note
Note: You should check to see if there are any Oracle DB's running first! If there are,
arrange for the DBA's to shut these down gracefully.
XSCF> console -d 0
Connect to DomainID 0?[y|n] :y
#● (hash dot)
XSCF> showdomainstatus -a
DID Domain Status
00 Booting/OpenBoot PROM prompt
01 -
XSCF> showboards -d 0
XSB DID(LSB) Assignment Pwr Conn Conf Test Fault
---- -------- ----------- ---- ---- ---- ------- --------
00-0 00(00) Assigned y y n Passed Normal
XSCF> poweroff -d 0
DomainIDs to power off:00
Continue? [y|n] :y
00 :Powering off
*Note*
This command only issues the instruction to power-off.
The result of the instruction can be checked by the "showlogs power".
To be able to send a break signal to the Domain, we must either turn the Panel Switch to the
Service position, which will of course require access to the Computer Hall, or we can turn the
Domain Secure Mode off. However, turning the Domain Secure Mode to off does have an
unwanted side effect in that it also disables the host watchdog reset.
Host Watchdog
Based on communication between XSCF and a domain, the host watchdog function checks
whether the domain is alive (heart beat or alive check). XSCF periodically monitors the
operational status of Solaris OS, to detect the Solaris OS hang-up. When detected the
Solaris OS hang-up, XSCF generates a Solaris OS panic on the relevant domain.
Therefore, to workaround this, we temporarily disable the Domain Secure Mode so that we can
send the break signal to the domain, and then enable it again as follows:
XSCF> showdomainmode -d 0
Host-ID :8525ded5
Diagnostic Level :min
Secure Mode :on
Autoboot :on
CPU Mode :auto
XSCF> sendbreak -d 0
Send break signal to DomainID 0?[y|n] :y
XSCF> console -d 0
Connect to DomainID 0?[y|n] :y
{22} ok #.
exit from console.
Note
On a console connection the key combination of "ctrl" and "\" ( CTRL- \ )will result in ascii value
0x1c (decimal = 28), which will also send a break signal to the domain.
XSCF> console -d 0
Connect to DomainID 0?[y|n] :y
c1718-3-92 console login: #.
Password: exit from console.
XSCF> showdomainstatus -d 0
DID Domain Status
00 Running
01 -
*Note*
This command only issues the instruction to reset.
The result of the instruction can be checked by the "showlogs power".
XSCF> console -d 0
Connect to DomainID 0?[y|n] :y
100% done: 94631 pages dumped, compression ratio 7.51, dump succeeded
rebooting...
Resetting... #.
exit from console.
XSCF> showdomainstatus -d 0
DID Domain Status
00 Powered Off
01 -
8. Administration
8.1. Platform
8.1.1. Configuring the XSCF
8.1.1.1. Login as the default User
To perform any configuration or administration on the XSCF, you need to login as a user with
administrative privileges. If you do not know the administrative user name or password, you may
need to login as the default user to setup one up or change the details.
By default, the XSCF comes with a couple of users configured (default and admin) which can be
used to login to the XSCF using the default account as follows:
• The best way to do this is to take your laptop into the Hall and connect directly to the back
of the Server
• You will need somebody in the Computer Hall standing next to the system during this
operation, you will also need a key for the Operator Panel (OPNL)
• The Cable you need is the same as you would use for a V210, V240, V440, V480 (9-pin
female ot RJ45 female adapter and a straight through ethernet cable)
• Plug the serial cable into the Serial port of the Active XSCF (the M4000 only has one
XSCF – check the XSCF description under the Architecture Section of this document)
• From your laptop, connect to the XSCF over the serial connection using any terminal
software (e.g Tera Term, HyperTerm or tip)
• Log in with the default user account. Follow the instructions to change the mode switch of
the operator panel. It will ask you to change the OPNL key position twice within one
minute, which is why you (or someone else) needs to be next to the Server.
For Example:
login: default
Change the panel mode switch to Service and press return...
Turn Key
Leave it in that position for at least 5 seconds. Change the panel mode
switch to Locked, and press return...
Turn Key
XSCF> showuser -l
User Name: xscf-admin
UID: 101
Status: Enabled
Minimum: 0
Maximum: 99999
Warning: 7
Inactive: -1
Last Change: Feb 25, 2009
Password Expires: Never
Password Inactive: Never
Account Expires: Never
Privileges: useradm
platadm
mode
fieldeng
auditadm
Note
If you are setting up an XSCF for the first time, it is possible to login across the Serial Port (through
a TC) as user admin. However, the admin user account is automatically disabled once someone
has either logged in as the default user or created a user account, so this user account will be of
little use once the Server is under live support:
XSCF> showuser -l
User Name: xscf-admin
UID: 101
Status: Enabled
Minimum: 0
Maximum: 99999
Warning: 7
Inactive: -1
Last Change: Feb 25, 2009
Password Expires: Never
Password Inactive: Never
Account Expires: Never
Privileges: useradm
platadm
mode
fieldeng
auditadm
However, you will need to apply all required priviliges in an additional step if you do it this way
as follows:
XSCF> setprivileges xscf-admin platadm useradm auditadm fieldeng mode
8.1.1.3. setautologout
The setautologout(8) command sets the session timeout time of the XSCF shell.
XSCF> showautologout
30min
XSCF> setautologout -s 60
60min
8.1.1.4. setupplatform
Setupplatform is used to configure the XSCF. The following example was taken from an M5000
XSCF> setupplatform
Do you want to set up an account? [y|n]: n
Do you want to set up networking? [y|n]: y
Do you want to set up the XSCF network interfaces? [y|n]: y
Do you want to configure xscf#0-lan#0? [y|n]: y
xscf#0-lan#0 ip address? []: 10.130.4.93
xscf#0-lan#0 netmask? []: 255.255.0.0
xscf#0-lan#0 default gateway? []: 10.130.0.1
xscf#0-lan#0 ip address: 10.130.4.93
xscf#0-lan#0 netmask: 255.255.0.0
xscf#0-lan#0 default gateway: 10.130.0.1
Are these settings correct? [y|n]: y
XSCF> setnetwork xscf#0-lan#0 -m 255.255.0.0 10.130.4.93
XSCF> setroute -c del -n 0.0.0.0 -m 0.0.0.0 -g 10.130.0.1 xscf#0-lan#0
XSCF> setroute -c add -n 0.0.0.0 -m 0.0.0.0 -g 10.130.0.1 xscf#0-lan#0
Do you want to configure xscf#0-lan#1? [y|n]: n
lan#0 is already configured:
lan#0 ip address: 10.130.4.93
lan#0 netmask: 255.255.0.0
Do you want to configure lan#0? [y|n]: n
Do you want to configure lan#1? [y|n]: n
DSCP network is already configured:
DSCP network: 192.168.224.0
DSCP netmask: 255.255.255.0
Do you want to set up the DSCP network? [y|n]: n
Domain name service is already configured:
Primary DNS server ip address:
Secondary DNS server ip address:
Tertiary DNS server ip address:
Domain name: ssclabs.net
XSCF#0 hostname: c1718-4-93-mgt-xscf0
Do you want to set up the domain name service? [y|n]: y
Primary DNS server ip address? []: 10.131.2.2
Do you want a secondary DNS server? [y|n]: n
Do you want to specify a domain name? [y|n]: y
Domain name []: ssclabs.net
XSCF#0 hostname []: c1718-4-92-mgt-xscf0
Primary DNS server ip address: 10.131.2.2
Secondary DNS server ip address:
Tertiary DNS server ip address:
Domain name: ssclabs.net
XSCF#0 hostname: c1718-4-92-mgt-xscf0
Are these settings correct? [y|n]: y
Do you want to set up the network time protocol? [y|n]: n
The ssh service is: enabled
Do you want to set up ssh? [y|n]: n
The https service is: disabled
Do you want to set up https? [y|n]: n
Do you want to configure email reports? [y|n]: n
Do you want to apply the network changes? [y|n]: n
The applynetwork command must be run in order for network changes to
take effect
Do you want to set up the chassis altitude? [y|n]: n
Do you want to set up the XSCF time zone? [y|n]: n
You can also run the setuplatform command with the “-p part” option which specifies the particular
setting you want, instead of doing them all as follows:
8.1.1.5. sethttps
The sethttps(8) command starts or stops the HTTPS service, which is used in the XSCF
network. Also, this command performs authentication-related settings for authentication used
in the HTTPS service.
It is arguable whether this protocol will be allowed through the firewalls on the NHS Project, but
it has been included in case it is ever required. The BUI Interface (Browser User Interface) is
extremely easy to use and gives you pretty much the same full functionality as the XSCF CLI.
The following example worked on servers in both Sun GMP and Linlithgow when configured by
members of the site based support teams and the Bridge during the Mx000 TOI's on XCP 1080.
XSCF> showhttps
HTTPS status: disabled
Server key: installed in Feb 18 17:20:43 GMT 2009
CA key: installed in Feb 18 17:20:43 GMT 2009
CA cert: installed in Feb 18 17:20:43 GMT 2009
CSR:
-----BEGIN CERTIFICATE REQUEST-----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-----END CERTIFICATE REQUEST-----
XSCF> rebootxscf
The XSCF will be reset. Continue? [y|n] :y
To use this, now fire up a web browser and point it to the following URL:
• https://XSCF_IP_Address
Note that you may need to add the XSCF IP Address under the “No Proxy For” settings of your
browser, otherwise the your proxy may no nothing about the IP address you are trying to connect
to.
8.1.1.6. restoredefaults
The restoredefaults(8) command deletes the setting and the log information that stored in the
server or the XSCF unit, and restores it to the state as of the factory shipment.
• xscfu Only restores XSCFU Log and OPNL Backup to factory default
• factory Restores all XSCF settings & OPNL to factory default
Note
You will need to physically power cycle the Server afterwards before it can be used. An example
of setting all XSCF settings & OPNL to factory default is as follows:
WARNING:
If this system does not have OPNL, this command will set all the user
settable XSCF configuration parameters to their default value as they
were set when the system was shipped out.
Furthemore, this command will delete all logs on both XSCFUs.
Check the man page of this command before you run it.
Continue?[yes/no](default no):yes
You must check the following points.
If you answer "yes" this command will HALT the XSCFU when it compltetes.
You will need to powercycle the system after the XSCF BOOT STOP.
Continue?[yes/no](default no):yes
The initialization of XSCFU will be started.
XSCFU : all data clear
OPNL : all data clear (exclude SYSTEM ID data)
XSCF will be automatically rebooted. Afterwards,
XSCFU will be initialized.
Continue?[yes/no](default no):yes
The NVRAM setting of XSCFU#0 was completed.
XSCF shutdown request was completed.
execute S10ioxoffXSCF> -- complete
Apr 15 11:07:50 FF2-24-0 XSCF[104]: XSCF shutdown sequence start
<snip>....XSCF reboot..<snip>
On the Data Center Servers (M8000/M9000), the replacefru command allows the Engineer to perform Hot
Replacement the CPU memory unit (CMU), I/O unit (IOU), FAN unit (FANU), Power Supply Unit (PSU),
XSCF unit (XSCFU) & DC to DC Converters (DDC - M8000 only).
Note that you must run replacefru before you remove any of these components, and in the case of the
CMU and IOU on the DC Servers, these components must be removed from the Domain before
replacement using either Static or Dynamic Reconfiguration (if supported)
The SAS Disks are Hot Plugable, but Solaris Administration (SVM/VxVM) will be required to ensure that
they are not in use when they are replaced.
There is also an LED locator light on the OPL Servers Operator Panel (OPNL) which you can turn on and
off manually from the XSCF to help locate a Server using the command setlocator.
8.1.2.1. showlocator
The showlocator(8) command displays the blink state of the CHECK LED on the operator panel,
one of the following states is displayed:
For example:
XSCF> showlocator
Locator LED status:Blinking
8.1.2.2. setlocator
The setlocator(8) command controls the blink state of the CHECK LED on the operator panel.
One of the following can be specified:
For example:
----------------------------------------------------------------------
Maintenance/Replacement Menu
Please select a FAN to be replaced.
----------------------------------------------------------------------
Maintenance/Replacement Menu
Status of the replaced FRU.
FRU Status
------------- --------
FAN_A#0 Normal
----------------------------------------------------------------------
The replacement of FAN_A#0 has completed normally.[f:finish] :f
----------------------------------------------------------------------
Maintenance/Replacement Menu
Please select a type of FRU to be replaced.
----------------------------------------------------------------------
Maintenance/Replacement Menu
Please select a PSU to be replaced.
Please confirm the Ready LED is not lit and that the Check LED is
blinking.
If this is the case, please replace PSU#0.
After replacement has been completed, please select[f:finish] :f
----------------------------------------------------------------------
Maintenance/Replacement Menu
Status of the replaced FRU.
FRU Status
------------- --------
PSU#0 Normal
----------------------------------------------------------------------
The replacement of PSU#0 has completed normally.[f:finish] :f
----------------------------------------------------------------------
Maintenance/Replacement Menu
Please select a type of FRU to be replaced.
Although an Upgrade can be performed while all Domains are up and running, each domain will need to
be shutdown and powered off and on again for the OBP and POST changes to take effect.
If you are upgrading the XSCFU F/W as part of a patch upgrade, perform the F/W Upgrade on the
XSCFU first and then patch the domains afterwards. This will avoid the need for the Domains to be
shutdown twice.
There are several key steps to performing an XSCF Firmware Upgrade as outined below. However, there
is a full example of a F/W Upgrade in the Appendix Section of this document.
You need to locate the correct F/W file for the target server. At the time of writing, XCP 1081 is the
latest version and therefore the required files are as follows:
These can either be download from the Sun Website below, or taken from the EIS DVD
• https://cds.sun.com/is-bin/INTERSHOP.enfinity/WFS/CDS-CDS_SMI-
Site/en_US/-/USD/ViewProductDetail-Start?ProductRef=OPL-M4-5-8-9000-
XCP-1080-SP-G-F@CDS-CDS_SMI
• Once the F/W tar.gz file has been copied to a Server or laptop, it can be loaded onto the XSCFU
from the XSCF command line (CLI) by the getflashimage command using FTP, HTTP or HTTPS or
through the Browser User Interface (BUI).
XSCF> getflashimage -l
Existing versions:
Version Size Date
FFXCP1072.tar.gz 49762997 Thu Mar 19 15:17:03 GMT 2009
Note that before it can be downloaded, the Server or laptop needs to be setup so that files can be
remotely copied using the desired protocol.
• After the XSCFU has rebooted, check that the updates have been applied and that the Current and
Reserve XCP are at the new version.
Allow 10 minutes after the XSCFU has rebooted for the changes to take effect as the XSCFU
Upgrade continues after the reboot. Example output can be found in the Appendix Section of this
document:
login: xscf-admin-user
password:
XSCF> version -c xcp
XSCF#0 (Active )
XCP0 (Current): 1072
XCP1 (Reserve): 1072
8.2. Domain
8.2.1. Create a Domain
8.2.1.1. Simple (Uni-XSB) Domain
Creating a Domain requires us to run several commands in order to completely configure and
assign all resources required to form a running domain as follows:
The M4000 only supports one domain when the PSB is in the Uni-XSB configuration. In this case,
all components of the single PSB (PSB0) are allocated to a single XSB (XSB 00-0), which in turn is
allocated to a single domain (DID 0).
XSCF> showdomainstatus -a
DID Domain Status
00 -
01 -
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Uni no
• Configure the XSB into Uni-mode (if required)
XSCF> setupfru -x 1 sb 0
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Uni no
XSCF> showboards -av
XSB R DID(LSB) Assignment Pwr Conn Conf Test Fault COD
---- - -------- ----------- ---- ---- ---- ------- -------- ----
00-0 SP Available n n n Unknown Normal n
XSCF> showdcl -v -d 0
DID LSB XSB Status No-Mem No-IO Float Cfg-policy
00 Powered Off FRU
00 00-0 False False False
01 -
02 -
03 -
04 -
05 -
06 -
07 -
08 -
09 -
10 -
11 -
12 -
13 -
14 -
15 -
• After you have setup the DCL's, you can assign the boards to the domain
XSCF> showdomainstatus -a
DID Domain Status
00 Powered Off
01 -
• Once the boards have been assigned to the domains, you can change some basic OBP
settings before powering it on:
XSCF> showdomainmode -d 0
Host-ID :8525ded4
Diagnostic Level :max
Secure Mode :on
Autoboot :on
CPU Mode :auto
XSCF> poweron -d 0
DomainIDs to power on:00
Continue? [y|n] :y
00 :Powering on
*Note*
This command only issues the instruction to power-on.
The result of the instruction can be checked by the "showlogs power".
XSCF> console -d 0
Connect to DomainID 0?[y|n] :y
POST Sequence 01 CPU Check
POST Sequence 02 Banner
LSB#05 (XSB#00-0): POST 2.7.0 (2008/09/25 13:56)
POST Sequence 03 Fatal Check
POST Sequence 04 CPU Register
POST Sequence 05 STICK
POST Sequence 06 MMU
POST Sequence 07 Memory Initialize
POST Sequence 08 Memory
POST Sequence 09 Raw UE In Cache
POST Sequence 0A Floating Point Unit
POST Sequence 0B SC
POST Sequence 0C Cacheable Instruction
POST Sequence 0D Softint
POST Sequence 0E CPU Cross Call
POST Sequence 0F CMU-CH
POST Sequence 10 PCI-CH
POST Sequence 11 Master Device
POST Sequence 12 DSCP
POST Sequence 13 SC Check Before STICK Diag
POST Sequence 14 STICK Stop
POST Sequence 15 STICK Start
POST Sequence 16 Error CPU Check
POST Sequence 17 System Configuration
POST Sequence 18 System Status Check
POST Sequence 19 System Status Check After Sync
POST Sequence 1A OpenBoot Start...
POST Sequence Complete.
{0} ok
However, in this example, we will split the fully loaded M4000 PSB into Quad-XSB configuration
and then create 2 domains. In this case, PSB (PSB0) will be split into 4 XSB's (XSB 00-0, 00-1,
00-2 & 00-3). XSB 00-0 will be assigned to Domain 0 (DID 0) and XSB 00-1 will be assigned to
Domain 1 (DID 1).
XSCF> showdomainstatus -a
DID Domain Status
00 -
01 -
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Uni no
XSCF> setupfru -x 4 sb 0
Operation has completed. However, a configuration error was detected.
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Quad no
Setup DID 0
XSCF> setdcl -d 0 -a 0=00-0
XSCF> showdcl -d 0 -v
DID LSB XSB Status No-Mem No-IO Float Cfg-policy
00 Powered Off FRU
00 00-0 False False False
01 -
02 -
03 -
04 -
05 -
06 -
07 -
08 -
09 -
10 -
11 -
12 -
13 -
14 -
15 -
XSCF> showboards -d 0 -v
XSB R DID(LSB) Assignment Pwr Conn Conf Test Fault COD
---- - -------- ----------- ---- ---- ---- ------- -------- ----
00-0 00(00) Assigned n n n Unknown Normal n
XSCF> showdomainmode -d 0
Host-ID :8525ded4
Diagnostic Level :max
Secure Mode :on
Autoboot :on
CPU Mode :auto
XSCF> poweron -d 0
DomainIDs to power on:00
Continue? [y|n] :y
00 :Powering on
*Note*
This command only issues the instruction to power-on.
The result of the instruction can be checked by the "showlogs power".
XSCF> console -d 0
Connect to DomainID 0?[y|n] :y
{0} ok
Setup DID 1
XSCF> setdcl -d 1 -a 0=00-1
XSCF> showdcl -d 1 -v
DID LSB XSB Status No-Mem No-IO Float Cfg-policy
01 Powered Off FRU
00 00-1 False False False
01 -
02 -
03 -
04 -
05 -
06 -
07 -
08 -
09 -
10 -
11 -
12 -
13 -
14 -
15 -
XSCF> showboards -d 1 -v
XSB R DID(LSB) Assignment Pwr Conn Conf Test Fault COD
---- - -------- ----------- ---- ---- ---- ------- -------- ----
00-1 * 01(00) Assigned n n n Unknown Normal n
XSCF> showdomainmode -d 1
Host-ID :8525ded4
Diagnostic Level :max
Secure Mode :on
Autoboot :on
CPU Mode :auto
XSCF> poweron -d 1
DomainIDs to power on:01
Continue? [y|n] :y
01 :Powering on
*Note*
This command only issues the instruction to power-on.
The result of the instruction can be checked by the "showlogs power".
XSCF> console -d 1
Connect to DomainID 1?[y|n] :y
{0} ok
In mirrored mode, as we saw in an earlier section,, both the the memory capacity and the the
interleave factor are halved. Memory Mirroring is configured by the Administrator from the
XSCF when running the setupfru command on each XSB. You can not configure Memory
Mirroring on the M3000 or on High End M8000/M9000 Servers on Quad-XSB's.
In the following example, we create a Simple Domain using a simple Uni-XSB but with Memory
Morroring enabled. Please note that to get the full benefit of memory mirroing, all XSB's which
are assigned to the domain need to have Memory Mirroring enabled.
XSCF> showdomainstatus -a
DID Domain Status
00 -
01 -
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Uni no
• Configuring the XSB into Uni-mode with Memory Mirroring enabled (Note the “-m y”
XSCF> setupfru -m y -x 1 sb 0
• Continue to create and start the Domain as before (Note Memory Mirror Mode = yes).
XSCF> showfru sb 0
Device Location XSB Mode Memory Mirror Mode
sb 00 Uni yes
XSCF> showdcl -d 0
DID LSB XSB Status
00 Powered Off
00 00-0
XSCF> showdomainstatus -a
DID Domain Status
00 Powered Off
01 -
XSCF> showdomainmode -d 0
XSCF> poweron -d 0
XSCF> console -d 0
This example shows how to remove a domain from a fully loaded M4000 with 1 running Domain using a
Uni-XSB.
XSCF> showdomainstatus -a
DID Domain Status
00 Running
01 -
XSCF> console -d 0
Connect to DomainID 0?[y|n] :y
XSCF> showdomainstatus -a
DID Domain Status
00 Booting/OpenBoot PROM prompt
01 -
XSCF> poweroff -d 0
DomainIDs to power off:00
Continue? [y|n] :y
00 :Powering off
*Note*
This command only issues the instruction to power-off.
The result of the instruction can be checked by the "showlogs power".
XSCF> showboards -d 0
XSB DID(LSB) Assignment Pwr Conn Conf Test Fault
---- -------- ----------- ---- ---- ---- ------- --------
00-0 00(00) Assigned n n n Passed Normal
XSCF> setdcl -d 0 -r 0 1
9. Dynamic Configuration
Dynamic Reconfiguration (DR) enables hardware resources such as processors, memory, and I/O to be added
and deleted even while the Solaris Operating System is running. DR has three basic functions;which are,
addition, deletion and move.
• Due to diagnostic requirements, the DR function works only on boards that have at least one CPU and
memory.
• For a DR attach to be successful, Solaris needs to be running on the Domain
• On domains running in SPARC64 VII Enhanced Mode (prtdiag will tell you), you cannot use DR to add a
system board that contains any SPARC64 VI processors. To add a SPARC64 VI processor you must
power off the domain, change it to SPARC64 VI Compatible Mode, then reboot the domain.
The project did, however, allow the addition of components using the following DR operations on OPL Servers:
9.1.1. XSCF
You can execute DR operations from the service processor by using the XSCF addboard command to
accomplish the following tasks:
• Board addition – Dynamically adds a system board to a domain or a new board to the system as
a whole
The DR operation to add a system board cannot be executed when the system board has only been
mounted. The DR operation is enabled by registering the system board in the DCL by using the XSCF
shell or XSCF Web. You must confirm that the system board to be added is registered in the DCL before
performing the DR
As well as adding system boards dynamically, DR also lets you order such reconfiguration to take place
the next time the affected domains are turned on or turned off, or the domain is rebooted. Use the
addboard command with the -c reserve option to specify this action.
9.1.2. Solaris
The cfgadm command can be run directly from the domain to get board status and to dynamically
configure PCI adapters only. The cfgadm command does not support DR operations for system boards.
9.2. DR Commands
XSCF shell commands for DR operations are classified into two types: DR display and DR operation commands.
A summary of the commands which can be used by Annuity during DR operations is as follows.
It is advisable to have a console session running to the Domain in a seperate window when perform a DR attach
operation so that any errors or issues, such as a domain hang, can be observed and reported.
In this example, we will assign an additional Uni-XSB System Board (XSB 01-0) to Domain 0 while Solaris is
running:
1. Login to XSCF.
Execute the showdcl(8) command to display domain information, and then check the operation
status of the domain. Based on the operation status of the domain, determine whether to perform the
DR operation or change the domain configuration.
XSCF> showdcl -d 0
DID LSB XSB Status
00 Running
00 00-0
01 01-0
Execute the showboards(8) command to display system board information, and then check the
status of the system board to be added and confirm its registration in the DCL.
XSCF> showboards -a
XSB DID(LSB) Assignment Pwr Conn Conf Test Fault
----------------------------------------------------------------
00-0 00(00) Assigned y y y Passed Normal
01-0 SP Available y n n Passed Normal
Execute the addboard(8) command to add the system board to the move-destination domain.
XSCF> addboard -c configure -d 0 01-0
When the addboard(8) command ends normally, execute the showdcl(8) command to check the
operation status of the domain, and then execute the showboards(8) command to check the status
of the added system board. If the addboard(8) command completes abnormally or leaves the board
in an unwanted status, investigate and escalate to SSC as appropriate:
XSCF> showdcl -d 0
DID LSB XSB Status
00 Running
00 00-0
01 01-0
XSCF> showboards -d 0
XSB DID(LSB) Assignment Pwr Conn Conf Test Fault
------------------------------------------------------------------
00-0 00(00) Assigned y y y Passed Normal
01-0 00(01) Assigned y y y Passed Normal
10. Housekeeping
10.1. Reporting
10.1.1. Configure SMTP
The mail reporting function used by the XSCF can send messages to the administrator and has the
following features:
• Notification by e-mail of any component faults in a server. Even if a system failure or a serious
error that disables reboot occurs, an e-mail message is sent.
• POP or SMTP authentication for sending of e-mail.
The setsmtp command can also be used without options. This will invoke an interactive mode that asks
you to provide:
XSCF> setemailreport
Enable Email Reporting? [no]:yes
Email Recipient Address :joe@10.6.15.50
Do you want to send a test mail now (Yes/No): no
10.2. Archiving
The persistent storage space on the service processor is limited. A portion of this space is set aside for logs,
such as audit logs and error logs. Due to the limited space, some logs can grow to the point where old log
entries must be overwritten or deleted to make space for new logs.
By default, log archiving is disabled. To use log archiving, you must configure an archive host, and then enable
log archiving on the service processor.
When enabled, log archiving periodically uses the secure copy program (scp) to transfer new log data to the
archive host. Log archiving also usessecure shell (ssh) to monitor the disk space consumed by archives. It
deletes older archives when necessary, so that the space consumed by the archives never exceeds the user-
configurable archive space limits.
For security reasons, log archiving does not automatically delete audit log archives. You can, however, still
manually delete audit log archives that are no longer needed.
10.2.1. setarchiving
The setarchiving command configures the log archiving. Setarchiving has many options, of which the
following are key:
The following example sets the archiving target system and password.
10.2.2. showarchiving
The showarchiving command displays the log archiving configuration and status.
XSCF> showarchiving
*** Archiving Configuration ***
Archiving state ---------- Disabled
Archive host ------------- Not configured
Archive directory -------- Not configured
User name for ssh login -- Not configured
Archive host fingerprint - Server authentication disabled
10.3. Auditing
An MX000 server logs all service processor events that could be relevant to security, such as system startup and
shutdown, user login and logout, and privilege changes. These events are tracked and implemented through:
• Audit records
• Audit trails
• Audit events
• Audit classes
• Audit policy
Because local space is limited to 4 megabytes, the partitions fill up quickly. If you do not configure the
audit policy to automatically transfer files to remote storage, you will have to intervene frequently or begin
to drop records. If you are unable to maintain consistent audit trails, the utility of the audit system is
limited. Typically, you either set up sufficient remote space and automatic transfers or disable the audit
capability.
• Audit records
• Audit trails
The default audit policy that is configured on the service processor is:
• Auditing is enabled.
• Records are dropped and counted when the audit trail is full.
• All events are enabled for auditing.
• The global user audit policy is enabled.
• Per-user audit policy for all users is enabled.
• Audit warning thresholds are set at 80 percent and 100 percent full.
• Email warnings are disabled.
10.3.6. setaudit
The setaudit command manages the collection of security-related data regarding the use of system
resources. This data can be used to assign responsibility for actions that have taken place on the system.
This command requires auditadm privileges to execute.
The following example changes the audit recording for various classes:
10.3.7. showaudit
The showaudit command displays the log archiving configuration and status. This command requires
auditadm, auditop, or escalation privileges to execute.
XSCF> showaudit
Auditing: enabled
10.3.8. viewaudit
The viewaudit command displays audit records. When invoked without options, viewaudit displays all
current local audit records. When invoked with options, viewaudit displays only the selected records. By
default, records are displayed in text format, one token per line, with a comma as the field separator. This
command requires auditadm or auditop privileges to execute.
The following example displays audit records for February 22, 2009:
The following example displays user audit records for a user named kevin:
The following example displays audit records between the dates of January 12, 2009 and January 14,
2009:
If the Service Processor is replaced, it will be auomatically comnfigured from the backup in the OPNL. If
there Service Processor has some pre-existing configuration and is not able to automatically confiigure
the Service Processor, it will display details to the serial port and request confirmation.
If the OPNL is replaced, it will automatically take a backup of the Service Processor once it has been
installed and the Platform powered on.
The XSCF configuration information can be saved in the device and can be restored using one of the
following two methods.
• The configuration information can be saved and restored when a USB device has been
connected to the USB connector mounted on the XSCF Unit front panel of the M4000/M5000/
M8000/M9000 servers or rear panel of the M3000 server.
• The configuration data is transmitted through the network with an encryption protocol.
Note
• The USB device should only be formatted using the FAT32 file system.
• Power off all domains when you restore the configuration information. Moreover, after
completing the command, turn the input power supply of the server off, then on.
10.4.2.1. dumpconfig
• Backup to a USB stick
10.4.2.2. restoreconfig
• Shutdown & Power off all domains.
• Turn the input power supply of the server off, then on.
11. Monitoring
The following section assumes you are running SunMC 3.6.1.
Sun Management Center 3.6.1 Version 2 add-on software introduces support for the following systems:
This add-on software for Sun SPARC Enterprise Mx000 servers provides the following features:
• Hardware monitoring
• Power management
• Domain management
• Dynamic reconfiguration
• FRU replacement (system board)
A combination of the following modules need to be download and installed on the SunMC Server, the SunMC
Console and the Mx000 Domains.
The XSCF requires no Add-on installation as the SunMC Agent is provided as part of the XCP. The XCP
contains the necessary firmware required to properly operate the XSCFU and the server components at a very
low level.
Plat Admin Module Platform administration module. Provides monitoring and active
SPARC Enterprise management capability for the Sun SPARC Enterprise Mx000 server.
Mx000 Resides on the Service Processor.
• https://cds.sun.com/is-bin/INTERSHOP.enfinity/WFS/CDS-CDS_SMI-
Site/en_US/-/USD/ViewProductDetail-Start?ProductRef=SunMC-3.6.1V3-G-
F@CDS-CDS_SMI
mx000-uksr: cd /export/downloads/sunmc3.6.1
mx000-uksr: cd disk2/image
mx000-uksr: /opt/SUNWsymon/sbin/es-inst -S .
----------------------------------------------------------------------------
This script installs Sun (TM) Management Center 3.6.1
----------------------------------------------------------------------------
XSCF> showsunmc
Agent Status: Disabled, stopped
Setup Status: Not set up
SunMC Server: NOT_SET
Security Seed: maplesyr
SNMPv1 Community String: public
Agent Port: 1161
Host Trap Port: 162
Host Event Port: 163
SNMP Agent Port: 161
Domain Agent Port: 1161
XSCF> showsunmc
Agent Status: Enabled, running
Setup Status: Set up
SunMC Server: 192.168.15.9
Security Seed: maplesyr
SNMPv1 Community String: public
Agent Port: 1161
Host Trap Port: 162
Host Event Port: 163
SNMP Agent Port: 161
Domain Agent Port: 1161
• It is not necessary to configure SNMP for SunMC to work, an example of an XSCFU which has not
has SNMP configured can be seen below:
XSCF> showsnmp
mx000-uksr: cd /export/downloads/sunmc3.6.1
mx000-uksr: cd disk2/image
mx000-uksr: /opt/SUNWsymon/sbin/es-inst -S .
----------------------------------------------------------------------------
This script installs Sun (TM) Management Center 3.6.1
----------------------------------------------------------------------------
You must perform setup before using Sun Management Center 3.6.1.
Do you want to run setup now (y|n|q) y
----------------------------------------------------------------------------
This script will help you to setup Sun (TM) Management Center 3.6.1.
----------------------------------------------------------------------------
Stopping platform component
Stopping agent component
Setting up Addon[s]...
---------------------------------------------------------------------
Starting Sun Management Center Mx000 Systems Domain Setup
---------------------------------------------------------------------
End of Installation
11.1.4. Configure the Mx000 XSCFU and Domains on the SunMC Server
Make sure that the Domains and the XSCFU are in the SunMC Server /etc/hosts file and that they are
pingable. We also need to ensure that ssh is working.
For each Domain, add a new entry with the following details
Hostname: Mx000-hostname
IP Address: <IP Address>
Port: 1161
12. Troubleshooting
This section contains information which should assist you to troubleshoot problems from the Domains and the Service
Processor. Note that there are additional examples in the Appendix Section of this document for reference.
12.1. Platform
12.1.1. snapshot
The snapshot(8) command provides a data-collection mechanism that enables rapid, reliable, and
flexible retrieval of diagnostic information on the Service Processor. Snapshot(8) collects the following
data: Configuration, Environmentals, Logs, Errors, and FRUID information. It transfers data to the
specified destination.
Basically, a snapshot is like an explorer, but run on the XSCFU. We provide it a target Server, Laptop or
USB Stick and it will generate based on the host name and IP address assigned to the Service
Processor and the UTC time (in hours, minutes, and seconds) and date on the Service Processor at
the time snapshot is invoked
(e.g jupiter_10.1.1.1_2006-07-08T22-33-44.zip)
The first mode is SSH target mode. The data collector runs in this mode when it is invoked with the
-t or -T option. In this mode, the data collector opens an SSH connection from the service
processor to th specified target (after appropriate authentication) and sends the zipped data
archive through the SSH connection to the target host. The transmission encryption in this mode is
provided by SSH.
The second mode is USB device mode. The data collector runs in this mode when it is invoked
with the -d flag. In this mode, the data collector’s output (which is the zip archive) is saved in a file
on a USB device. No transmission encryption occurs in this mode because the data stays local to
the service processor.
12.1.1.3. Examples
• The following example downloads the snapshot using a public key through SSH and specifies
the remote host:
• The following example downloads the snapshot using a public key through SSH and specifies
the remote host:
XSCF> snapshot -T -k download
Downloading Public Key from ‘10.6.15.50’...
Key fingerprint in md5: c9:e0:bc+b2:1a:80:29:24:13:d9:f1:13:f5:5c:2c:0f
Accept this public key (yes/no)? Y
Enter ssh password for user ‘matt’ on host ‘10.6.15.50’
Setting up ssh connection to remote host...
Collecting data into matt@godzilla:/export/home/matt/logs/kongsp0_
10.6.15.41_2008-02-25T06-21-17.zip
Data collection complete.
• The following example logs files only (no command output), using no public key:
XSCF> snapshot -t gary@10.6.15.50:/export/home/gary/logs -k none -l
Enter ssh password for user ’gary’ on host ’10.6.15.50’
Log only mode. No commands will be collected.
Setting up ssh connection to remote host...
Collecting data into gary@10.6.15.50:/export/home/gary/logs/kongsp0_
10.6.15.41_2008-02-25T06-24-57.zip
Data collection complete.
Change:
COMMAND="snapshot -T -D /tmp"
To:
COMMAND="snapshot -y -T -D /export/Mx000/snaphot"
Note – The xscfextended option to Sun Explorer assumes that archiving is set up and works.
XSCF> showstatus
No failures found in System Initialization.
12.1.2.2. showlogs
The showlogs command displays the contents of the specified log file. The output is displayed
using a timestamp, starting with the oldest date. The showlogs command can be run with a
combination of options (-d DID, -M, -v, -V)
12.1.2.8. showmonitorlog
showmonitorlog is like 'showlogs monitor', but it effictively performs a 'tail -f' on the monitor log
XSCF> showmonitorlog
• Error Log
• Fault Log
For example, here we have the fdump output from a M5000 with a failed fan in a PSU (a more
detailed example of this failure can be found a few pages further on in this User Guide):
XSCF> fmdump
TIME UUID MSG-ID
Feb 09 15:20:46.7179 88f2ccb8-cf7d-4dae-ba65-4ffc1f655124 SCF-8001-KC
Feb 09 15:30:46.6667 554cc530-a17a-4d76-ac56-2685c1c97daf SCF-8002-FA
Feb 24 02:14:03.0020 bf74be0f-fea6-4299-8c04-7688870ea64b SCF-8005-YH
Feb 24 02:14:04.4976 7f2f29b1-811f-403c-a058-6088c5c0969b SCF-8005-YH
XSCF> fmdump -v -m
MSG-ID: SCF-8001-KC, TYPE: upset, VER: 1, SEVERITY: Critical
EVENT-TIME: Mon Feb 9 15:20:46 GMT 2009
PLATFORM: SPARC Enterprise M5000, CSN: BCF0904031, HOSTNAME: c1718-4-92-mgt-xscf0
SOURCE: sde, REV: 1.16
EVENT-ID: 88f2ccb8-cf7d-4dae-ba65-4ffc1f655124
DESC: Failure to boot Solaris on a domain via an internal disk drive on the SAS
controller.
Refer to http://www.sun.com/msg/SCF-8001-KC for more information.
AUTO-RESPONSE: Nothing is deconfigured.
IMPACT: The domain drops to the OpenBoot PROM 'ok' prompt.
REC-ACTION: The platform administrator should investigate the cause of the boot problem
and schedule a service action, if necessary.
Please consult the detail section of the knowledge article for additional information.
12.1.3.2. fmadm
The fmadm command can display the configuration of the Fault Manager itself, including the
module name, version, and description of each component module. Fault Manager modules
provide services such as automated diagnosis, self healing, and messaging for hardware and
software present on the system.
Certain functions of fmadm (e.g repair & unload) are only available if you request a password to go
into service mode (see service(8)), however, you can run 'fmadm config' (below)
12.1.3.3. fmstat
The fmstat utility can be used to report statistics associated with the Fault Manager (fault manager
daemon), and its associated set of modules. You can use fmstat to view statistics for diagnosis
engines and agents that are currently participating in fault management
XSCF> fmstat
module ev_recv ev_acpt wait svc_t %w %b open solve memsz bufsz
case-close 14 0 0.0 3.1 0 0 0 0 686b 0
emailprv 14 0 0.0 7.5 0 0 0 0 0 0
event-transport 0 0 0.0 1.8 0 0 0 0 5.8K 0
faultevent-post 14 0 0.0 113.5 0 0 0 0 0 0
flush 28 0 0.0 0.2 0 0 0 0 0 0
fmd-self-diagnosis 0 0 0.0 1.6 0 0 0 0 0 0
iox_agent 0 0 0.0 1.6 0 0 0 0 0 0
reagent 0 0 0.0 1.6 0 0 0 0 0 0
sde 14 14 0.0 81.8 0 0 0 14 159K 0
snmp-trapgen 14 0 0.0 4.6 0 0 0 0 0 0
sysevent-transport 0 0 0.0 68.4 0 0 0 0 0 0
syslog-msgs 14 0 0.0 11.4 0 0 0 0 97b 0
12.1.4. Diagnostics
12.1.4.1. testsb
The testsb command performs an initial diagnosis of the specified PSB, you must have
platadm or fieldeng privileges to run testsb
The configuration of the PSB and operation of each device mounted on the PSB are checked.
After the diagnostics, the result is displayed. The PSB must not be configured in the domain, or the
domain in which the PSB configured must be powered off. The result also can be seen in "Test"
and "Fault" displayed by the showboards command.
In the following example, we have run testsb against PSB#0 on an M4000. Note that Domain 0 to
which XSB#0 is assigned is powered off:
XSCF> showdomainstatus -a
DID Domain Status
00 Powered Off
01 -
XSCF> testsb 0
Initial diagnosis is about to start, Continue?[y|n] :y
SB#00 power on sequence started.
0..... 30.....end
Initial diagnosis started. [1800sec]
0..... 30..... 60...end
Initial diagnosis has completed.
SB power off sequence started. [1200sec]
0.end
SB powered off.
XSB Test Fault
---- ------- --------
00-0 Passed Normal
00-1 Passed Normal
• Service mode
• Escalation mode
1. Obtain a valid password from an authorized service provider to execute the appropriate
enable command. You will be required to supply the serial number of the server as well as
the XSCF version number.
• Service Mode passwords can be generated form the following Sun Internal
Website:
• http://oplpass.sfbay
• Escalation Mode passwords can only be generated by PTS via an SSC Logged
call:
2. Enable the requested mode of operation using the appropriate enable command,
enableservice or enableescalation.
4. Use the mode password that is generated by the enable command to enter into the
appropriate mode, service or escalation.
Note – The service mode is intended to be used only by authorized personnel. The escalation mode is
only used during an escalation and with the guidance of engineering personnel.
2. Get the required information (XSCF Version and Serial No.) from the Server
XSCF> showhardconf -M
SPARC Enterprise M4000;
+ Serial:BCF071100U; Operator_Panel_Switch:Service;
It will ask for your SWAN login and Password, then display a Password Generation Grid
like the one below:
MODE : service
VERSION : 01.08.0003
SERIAL : BCF071100U
UTC DATE : 01/07/2009 09:50
XSCF> service
Password: WET RUDE LILY
service> showmodes
Current mode: service
Enabled for: service
Expires: Fri Jan 9 10:06:11 2009
service> exit
In addition to the additional options and arguments to exisiting commands, Service Mode offers
the following additional commands on XCP 1081:
• clearfault
• confxscf
• enablecodboard
• erasenvram
• setdumphost
• showdumphost
• shownvram
XSCF> escalation
Password: LOVE CAT NAPS
escalation> showmodes
Current mode: escalation
Enabled for: escalation
Expires: Fri Jan 9 10:06:11 2009
escalation> exit
12.2. Domain
12.2.1. DSCP
You can check that the Service Processor is able to communicate with the OPL Domains through the
DSCP Network as follows:
# ifconfig -a
<... stuff deleted ...>
sppp0: flags=10010008d1<UP,POINTOPOINT,RUNNING,NOARP,MULTICAST,IPv4,FIXEDMTU> mtu
1500 index 4
inet 192.168.224.2 --> 192.168.224.1 netmask ffffff00
ether 0
If there is no sppp0 interface configured, check to see if the required Services, specifically dscp, are
running as follows:
# svcs dscp
STATE STIME FMRI
online 14:40:37 svc:/platform/sun4u/dscp:default
# svcs dcs
STATE STIME FMRI
online 14:40:37 svc:/platform/sun4u/dcs:default
# svcs sckmd
STATE STIME FMRI
online 14:40:29 svc:/platform/sun4u/sckmd:default
If they are not running, check to see if there are any problems
# svcs -l dscp
fmri svc:/platform/sun4u/dscp:default
name DSCP Service
enabled true
state online
next_state none
state_time 12 March 2009 14:40:37 GMT
logfile /var/svc/log/platform-sun4u-dscp:default.log
restarter svc:/system/svc/restarter:default
contract_id 44
dependency require_any/error svc:/milestone/network (online)
dependency require_all/none svc:/system/filesystem/local (online)
dependency optional_all/none svc:/platform/sun4u/sckmd:default (online)
# tail /var/svc/log/platform-sun4u-dscp:default.log
[ Feb 27 16:45:07 Executing start method ("/lib/svc/method/svc-dscp start") ]
[ Feb 27 16:45:12 Method "start" exited with status 0 ]
[ Feb 27 16:55:38 Executing start method ("/lib/svc/method/svc-dscp start") ]
[ Feb 27 16:55:44 Method "start" exited with status 0 ]
[ Feb 27 16:59:00 Executing start method ("/lib/svc/method/svc-dscp start") ]
[ Feb 27 16:59:06 Method "start" exited with status 0 ]
[ Mar 11 15:28:17 Executing start method ("/lib/svc/method/svc-dscp start") ]
[ Mar 11 15:28:23 Method "start" exited with status 0 ]
[ Mar 12 14:40:32 Executing start method ("/lib/svc/method/svc-dscp start") ]
[ Mar 12 14:40:37 Method "start" exited with status 0 ]
12.2.2. SunVTS
SunVTS, Sun Validation Test Suite, tests and validates Sun hardware by verifying the connectivity
and functionality of hardware devices, controllers and peripherals
It is better to do this over a remote network login rather than on the console as SunVTS output will fill
up the SMS console logs
• Disable rpcbind
13. Appendix
13.1. PSU Failure Example
The following example is of a PSU fan failure on an M5000. It is used to illustrate the use of the XSCFU
commands to identify and view information about a fault on the system:
XSCF> replacefru
----------------------------------------------------------------------
Maintenance/Replacement Menu
Please select a type of FRU to be replaced.
----------------------------------------------------------------------
Maintenance/Replacement Menu
Please select a PSU to be replaced.
Please confirm the Ready LED is not lit and that the Check LED is
blinking.
If this is the case, please replace PSU#0.
After replacement has been completed, please select[f:finish] :f
[Warning:021]
Diagnostics have detected a hardware error.
This FRU cannot be used.
Please check the FRU.
Replacement operation is terminated.[c:cancel] :c
----------------------------------------------------------------------
Maintenance/Replacement Menu
Please select a type of FRU to be replaced.
XSCF> showstatus
* PSU#0 Status:Faulted;
XSCF>
XSCF> fmdump
TIME UUID MSG-ID
Feb 09 15:20:46.7179 88f2ccb8-cf7d-4dae-ba65-4ffc1f655124 SCF-8001-KC
Feb 09 15:30:46.6667 554cc530-a17a-4d76-ac56-2685c1c97daf SCF-8002-FA
Feb 24 02:14:03.0020 bf74be0f-fea6-4299-8c04-7688870ea64b SCF-8005-YH
Feb 24 02:14:04.4976 7f2f29b1-811f-403c-a058-6088c5c0969b SCF-8005-YH
XSCF> showhardconf
SPARC Enterprise M5000;
+ Serial:BCF0904031; Operator_Panel_Switch:Locked;
+ Power_Supply_System:Single; SCF-ID:XSCF#0;
+ System_Power:On; System_Phase:Cabinet Power On;
Domain#0 Domain_Status:Running;
XSCF> getflashimage -l
Existing versions:
Version Size Date
FFXCP1072.tar.gz 49762997 Thu Mar 19 15:17:03 GMT 2009
login: Mar 19 15:32:42 yama XSCF flashupdate[699]: XSCF download has been
completed (XSCFU=0,bank=1,XCP version=1072:last version=1081, Firmware
Element ID=00:version=01070001:last version=01080001)
Mar 19 15:32:42 yama XSCF flashupdate[699]: XSCF download is started
(XSCFU=0,bank=1,XCP version=1072:last version=1081, Firmware Element
ID=01:version=01070005:last version=01080004)
Mar 19 15:35:16 yama XSCF flashupdate[699]: XSCF download has been
completed (XSCFU=0,bank=1,XCP version=1072:last version=1081, Firmware
Element ID=01:version=01070005:last version=01080004)
Mar 19 15:35:16 yama XSCF flashupdate[699]: XSCF download is started
(XSCFU=0,bank=1,XCP version=1072:last version=1081, Firmware Element
ID=02:version=01070000:last version=01080000)
Mar 19 15:35:42 yama XSCF flashupdate[699]: XSCF download has been
completed (XSCFU=0,bank=1,XCP version=1072:last version=1081, Firmware
Element ID=02:version=01070000:last version=01080000)
Mar 19 15:35:42 yama XSCF flashupdate[699]: XSCF download is started
(XSCFU=0,bank=1,XCP version=1072:last version=1081, Firmware Element
ID=03:version=01070001:last version=01080004)
Mar 19 15:37:35 yama XSCF flashupdate[699]: XSCF download has been
completed (XSCFU=0,bank=1,XCP version=1072:last version=1081, Firmware
Element ID=03:version=01070001:last version=01080004)
Mar 19 15:37:35 yama XSCF flashupdate[699]: XSCF download is started
(XSCFU=0,bank=1,XCP version=1072:last version=1081, Firmware Element
ID=04:version=01070005:last version=01080004)
Mar 19 15:38:51 yama XSCF flashupdate[699]: XSCF download has been
login: xscf-admin-user
Password:
XSCF> version -c xcp
XSCF#0 (Active )
XCP0 (Current): 1072
XCP1 (Reserve): 1072
mx000-uksr: cd disk2/image
mx000-uksr: /opt/SUNWsymon/sbin/es-inst -S .
----------------------------------------------------------------------------
This script installs Sun (TM) Management Center 3.6.1
----------------------------------------------------------------------------
## Installing part 1 of 1.
/opt/SUNWsymon/base/lib/sparc-sun-solaris2.10/libdr.so <symbolic link>
/opt/SUNWsymon/base/lib/sparc-sun-solaris2.10/libdr.so.1.0
/opt/SUNWsymon/base/lib/sparc-sun-solaris2.10/pkgdr.so <symbolic link>
/opt/SUNWsymon/base/lib/sparc-sun-solaris2.10/pkgdr.so.1.0
/opt/SUNWsymon/modules/cfg/dr-opl-d.flt
/opt/SUNWsymon/modules/cfg/dr-opl-d.x
/opt/SUNWsymon/modules/cfg/dr-opl-daps-images-j.x
/opt/SUNWsymon/modules/cfg/dr-opl-disk-images-j.x
/opt/SUNWsymon/modules/cfg/dr-opl-emptyslots-images-j.x
/opt/SUNWsymon/modules/cfg/dr-opl-family-j.x
/opt/SUNWsymon/modules/cfg/dr-opl-io-images-j.x
/opt/SUNWsymon/modules/cfg/dr-opl-m.x
/opt/SUNWsymon/modules/cfg/dr-opl-models-d.x
/opt/SUNWsymon/modules/cfg/dr-opl-pci-images-j.x
/opt/SUNWsymon/modules/cfg/dr-opl-scsi-images-j.x
/opt/SUNWsymon/modules/cfg/dr-opl-slots-images-j.x
/opt/SUNWsymon/modules/cfg/dr-shell.tcl
/opt/SUNWsymon/modules/sbin/dr-opl-availability.sh
/opt/SUNWsymon/modules/sbin/execute-dapsdr-cmd.sh
/opt/SUNWsymon/modules/sbin/execute-sbddr-cmd.sh
/opt/SUNWsymon/modules/sbin/finish-sbddr-cmd.sh
## Installing part 1 of 1.
/opt/SUNWsymon/addons/OplDomAdmin/classes/SetupArchive.class
/opt/SUNWsymon/addons/OplDomAdmin/lib/locale/install/classes/OplDomAdmin.properties
/opt/SUNWsymon/addons/OplDomAdmin/sbin/addon_common.sh
/opt/SUNWsymon/addons/OplDomAdmin/sbin/es-setup-esoplda.sh
/opt/SUNWsymon/addons/OplDomAdmin/sbin/es-setup.sh
/opt/SUNWsymon/addons/OplDomAdmin/sbin/es-uninst-esoplda.sh
/opt/SUNWsymon/addons/OplDomAdmin/sbin/es-uninst.sh
/opt/SUNWsymon/addons/OplDomAdmin/sbin/opl-dmn-update-cfg.sh
/opt/SUNWsymon/base/cfg/base-Config-ReaderOpl-D.tcl
/opt/SUNWsymon/base/lib/sparc-sun-solaris2.10/libconfigdopl.so <symbolic link>
/opt/SUNWsymon/base/lib/sparc-sun-solaris2.10/libconfigdopl.so.1.0
/opt/SUNWsymon/base/lib/sparc-sun-solaris2.10/pkgconfigdopl.so <symbolic link>
/opt/SUNWsymon/base/lib/sparc-sun-solaris2.10/pkgconfigdopl.so.1.0
/opt/SUNWsymon/bin/configdopld
/opt/SUNWsymon/bin/configdopld.bin
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-d.def
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-d.flt
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-d.prc
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-d.rul
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-d.x
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-m.x
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-models-d.x
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-ruleinit-d.x
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl-ruletext-d.x
/opt/SUNWsymon/modules/cfg/Config-Reader-Opl.schema
/opt/SUNWsymon/modules/sbin/Config-Reader-Opl-schema.sh
/opt/SUNWsymon/modules/sbin/Config-Reader-Opl-shell.sh
[ verifying class <none> ]
Copyright 2007 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
Sun MC DomMonit SPARC Enterprise Mx000 Server and Agent Support(sparc) 3.6.1-
v3,REV=2.10.2007.07.11
Using </opt> as the package base directory.
## Processing package information.
## Processing system information.
5 package pathnames are already properly installed.
Installing Sun MC DomMonit SPARC Enterprise Mx000 Server and Agent Support as
<SUNWesopldi>
## Installing part 1 of 1.
[ verifying class <none> ]
Copyright 2007 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
French Sun Management Center Agent layer support for OPL domains(all) 3.6.1-
Installing French Sun Management Center Agent layer support for OPL domains as
<SUNWfplda>
## Installing part 1 of 1.
/opt/SUNWsymon/addons/OplDomAdmin/lib/locale/install/classes/OplDomAdmin_fr.properties
/opt/SUNWsymon/lib/locale/fr/LC_MESSAGES/SUNWesoplda.mo
/opt/SUNWsymon/lib/locale/fr_FR.ISO8859-15/LC_MESSAGES/SUNWesoplda.mo <symbolic link>
[ verifying class <none> ]
Copyright 2007 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
Japanese Sun Management Center Agent layer support for OPL domains(all) 3.6.1-
v3,REV=2007.07.23.17.12.17
Using </opt> as the package base directory.
## Processing package information.
## Processing system information.
11 package pathnames are already properly installed.
Installing Japanese Sun Management Center Agent layer support for OPL domains as
<SUNWjplda>
## Installing part 1 of 1.
/opt/SUNWsymon/addons/OplDomAdmin/lib/locale/install/classes/OplDomAdmin_ja.properties
/opt/SUNWsymon/lib/locale/ja/LC_MESSAGES/SUNWesoplda.mo
[ verifying class <none> ]
Copyright 2007 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
Korean Sun Management Center Agent layer support for OPL domains(all) 3.6.1-
v3,REV=2007.07.23.17.12.46
Using </opt> as the package base directory.
## Processing package information.
## Processing system information.
11 package pathnames are already properly installed.
Installing Korean Sun Management Center Agent layer support for OPL domains as
<SUNWkplda>
## Installing part 1 of 1.
/opt/SUNWsymon/addons/OplDomAdmin/lib/locale/install/classes/OplDomAdmin_ko.properties
/opt/SUNWsymon/lib/locale/ko/LC_MESSAGES/SUNWesoplda.mo
[ verifying class <none> ]
Copyright 2007 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
Simplified Chinese Sun Management Center Agent layer support for OPL domains(all) 3.6.1-
v3,REV=2007.07.23.17.12.26
Using </opt> as the package base directory.
## Processing package information.
## Processing system information.
11 package pathnames are already properly installed.
Installing Simplified Chinese Sun Management Center Agent layer support for OPL domains
as <SUNWcplda>
## Installing part 1 of 1.
Traditional Chinese Sun Management Center Agent layer support for OPL domains(all)
3.6.1-v3,REV=2007.07.23.17.12.55
Using </opt> as the package base directory.
## Processing package information.
## Processing system information.
11 package pathnames are already properly installed.
Installing Traditional Chinese Sun Management Center Agent layer support for OPL domains
as <SUNWhplda>
## Installing part 1 of 1.
/opt/SUNWsymon/addons/OplDomAdmin/lib/local/install/classes/OplDomAdmin_zh_TW.properties
/opt/SUNWsymon/lib/locale/zh_TW/LC_MESSAGES/SUNWesoplda.mo
[ verifying class <none> ]
Copyright 2007 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
You must perform setup before using Sun Management Center 3.6.1.
Do you want to run setup now (y|n|q) y
----------------------------------------------------------------------------
This script will help you to setup Sun (TM) Management Center 3.6.1.
----------------------------------------------------------------------------
Stopping platform component
Stopping agent component
Setting up Addon[s]...
---------------------------------------------------------------------
Starting Sun Management Center Mx000 Systems Domain Setup
---------------------------------------------------------------------
Additional Setup information is available in the
Sun Management Center 3.6.1 Version 2 Add-On Software Supplement for SPARC Enterprise
Servers.
End of Installation
13.4. Bookmarks
13.4.1. Documentation
• docs.sun.com M3000
• http://docs.sun.com/app/docs/prod/sparc.m3k?l=en&a=view
• docs.sun.com M4000
• http://docs.sun.com/app/docs/prod/sparc.m4k?l=en&a=view
• docs.sun.com M5000
• http://docs.sun.com/app/docs/prod/sparc.m5k?l=en&a=view
• docs.sun.com M8000
• http://docs.sun.com/app/docs/prod/sparc.m8k?l=en&a=view
• docs.sun.com M9000
• http://docs.sun.com/app/docs/prod/sparc.m9k?l=en&a=view
13.4.2. Support
• Onestop M3000
• https://sunspace.sfbay.sun.com/display/Onestop/SPARC%20Enterprise%20M3000
• Onestop M4000
• https://sunspace.sfbay.sun.com/display/Onestop/SPARC%20Enterprise%20M4000
• Onestop M5000
• https://sunspace.sfbay.sun.com/display/Onestop/SPARC%20Enterprise%20M5000
• Onestop M8000
• https://sunspace.sfbay.sun.com/display/Onestop/SPARC%20Enterprise%20M8000
• Onestop M9000
• https://sunspace.sfbay.sun.com/display/Onestop/SPARC%20Enterprise%20M9000
• SSE M3000
• http://sunsolve.central.sun.com/handbook_internal/Systems/SE_M3000/SE_M3000.html
• SSE M4000
• http://sunsolve.central.sun.com/handbook_internal/Systems/SE_M4000/SE_M4000.html
• SSE M5000
• http://sunsolve.central.sun.com/handbook_internal/Systems/SE_M5000/SE_M5000.html
• SSE M8000
• http://sunsolve.central.sun.com/handbook_internal/Systems/SE_M8000/SE_M8000.html
• SSE M9000
• http://sunsolve.central.sun.com/handbook_internal/Systems/SE_M9000/SE_M9000.html
13.4.3. Training
• WZT-4110 Sun SPARC Enterprise MX000 Server Overview, Architecture, and Components
• https://learning.sun.com/display/courses/WZT-
4110+Sun+SPARC+Enterprise+MX000+Server+Overview%2C+Architecture
%2C+and+Components
• WGS-PREX-4110 Sun SPARC Enterprise MX000 Server Overview and Architecture Assessment
• https://learning.sun.com/display/courses/WZT-
4110+Sun+SPARC+Enterprise+MX000+Server+Overview%2C+Architecture
%2C+and+Components
• WXI-4110 Sun SPARC Enterprise MX000 Server Support Readiness Training Assessment
• https://learning.sun.com/display/courses/WXI-
4110+Sun+SPARC+Enterprise+MX000+Server+Support+Readiness+Training+Assessment
13.4.4. Sales
• sun.com M3000
• http://www.sun.com/servers/midrange/m3000/index.xml
• sun.com M4000
• http://www.sun.com/servers/midrange/m4000/index.xml
• sun.com M5000
• http://www.sun.com/servers/midrange/m5000/index.xml
• sun.com M8000
• http://www.sun.com/servers/highend/m8000/index.xml
• sun.com M9000
• http://www.sun.com/servers/highend/m9000/index.xml
13.4.5. Various
• Sun Management Center 3.6.1 Version 3 Add-On Software
• https://cds.sun.com/is-bin/INTERSHOP.enfinity/WFS/CDS-CDS_SMI-
Site/en_US/-/USD/ViewProductDetail-Start?ProductRef=SunMC-3.6.1V3-G-F@CDS-
CDS_SMI
• SUN SPARC® ENTERPRISE M4000, M5000, M8000, AND M9000 SERVER ARCHITECTURE
• http://www.sun.com/servers/sparcenterprise/SPARCEnt-Arch-Final.pdf
• EIS Checklists
• http://www.eis.central/checklists/checklists.html
• Sun Doc 235101 - Sun[TM] Sparc Enterprise Mx000 (OPL) Servers Memory Configuration Rules
• http://sunsolve.sun.com/search/document.do?assetkey=1-61-235101-
1&searchclause=235101
• Mx000 Seies Resources – Excellent Internal Site! (includes “Deep Dive” Presentation)
• http://esp.west/~grc/opl/
• Sun TSC Internal Home Page for Mx000 Servers (great startting point for any Technical Information)
• http://panacea.central.sun.com/twiki/bin/view/Products/ProdInfoSunSPARCEnterpriseMx000