SGI® Octane™ III Deskside Cluster Server User’s Guide Document Number 007-5633-001
COPYRIGHT © 2010 SGI. All rights reserved; provided portions may be copyright in third parties, as indicated elsewhere herein. No permission is granted to copy, distribute, or create derivative works from the contents of this electronic documentation in any manner, in whole or in part, without the prior written permission of SGI.
Record of Revision Version Description -001 March, 2010 First release 007-5633-001 iii
Contents List of Figures . 1. . . . . . . . . . . . . . . . . . . . vii System Quick Start and Overview . . . . . . . . . . . . . . . . . . 1 ESD and Safety Precautions. . . . . . . . . . . . . . . . . . . . . 1 Quick Setup Procedures . . . . . . . . . . . . . . . . . . . . . . 2 Unpacking the System . . . . . . . . . . . . . . . . . . . . . 2 Choosing a Site . . . . . . . . . . . . . .
Contents 2. System Operation and Troubleshooting . . . . . . . . . . . . . . . . . 17 Operating Location Considerations . Server Precautions . 3. 4. . . . . . . . . . . . . . . . . 17 . . . . . . . . . . . . . . . . 18 . . . . Troubleshooting the System . . . . . . . . . . . . . . . . . . . . . 18 Enclosure Power Supply Status LEDs . . . . . . . . . . . . . . . . . 19 System Fan Failure . . . . . . . . .
List of Figures 007-5633-001 Figure 1-1 Octane III Unpacking Example . . . . . . . . . . . . 3 Figure 1-2 Octane III Rear Panel Connectors . . . . . . . . . . . 5 Figure 1-3 Pushing the Enable Power Switch . . . . . . . . . . . 6 Figure 1-4 SGI Octane III System Example Front View . . . . . . . . 8 Figure 1-5 Octane III Fan Locations . . . . . . . . . 12 Figure 1-6 Example Functional Block Diagram of an Octane III Server Board. . .
Chapter 1 1. System Quick Start and Overview This chapter provides quick setup and overview information for your SGI Octane III cluster server. Operating precautions are provided in this chapter, followed by basic setup instructions and a general overview of the product. Before starting and operating your system, familiarize yourself with the safety information in the following section: ESD and Safety Precautions Caution: Observe all ESD precautions.
1: System Quick Start and Overview Quick Setup Procedures Use the information in the following subsections to get your Octane III unpacked, setup and working in your office or lab environment. If your system ships with hardware, software or media not listed in this guide, please see its accompanying documentation for installation instructions. Unpacking the System Open the SGI Octane III system packaging container and inspect all shipped items for evidence of mishandling during transit.
Quick Setup Procedures Save the boxes and packing materials and reverse the previous steps if repacking is needed.
1: System Quick Start and Overview Choosing a Site The Silicon Graphics Octane III cluster server operates reliably within normal air conditioned office or lab environments. Select a site that meets these criteria: • Near two properly earthed, grounded, three-pronged power outlets, as follows: – In the U.S. and Canada: two single-phase NEMA 5-15R outlets for 100-120 V. Connectors will differ for 200-240 V applications.
Quick Setup Procedures System Connections and Power On Once you have located the system in its operating environment, use the following steps and illustrations to connect to power, Ethernet, monitor/keyboard/mouse and optional components: 1. Connect the keyboard and mouse to USB connectors on the rear of the enclosure. 2. Connect your monitor to the VGA or optional DVI video connector on the rear panel of the system. See Table 1-1 on page 9 for more information on video/optional connectors. 3.
1: System Quick Start and Overview The system starts and displays a welcome message. After that, a series of power-on self-test (POST) messages appear. The POST messages indicate whether or not the system is running correctly. If the system does not turn on or boot after pressing the enable power switch, see the next section for possible causes.
System Overview • No power comes from the grounded power outlet. Ask an electrician to check your power outlet. If the system still fails to boot, contact your SGI service representative for assistance. In North America, call 1-800-800-4SGI; international customers can visit http://www.sgi.com/support/ Click on the “Support Centers” link under the “Online Support” heading for information on how to contact your nearest SGI customer support center.
1: System Quick Start and Overview Latch to open cover Figure 1-4 8 SGI Octane III System Example Front View 007-5633-001
System Overview I/O Panel Overview Table 1-1 lists the standard and optional connectors used with the Octane III cluster server products. All the system connectors are located on the rear of the enclosure. Table 1-1 Octane III Cluster Server External Ports Available Number of ports Type of port Internal port connection Connector used Four USB 2.
1: System Quick Start and Overview Physical Specifications Table 1-2 shows the physical specifications for the Silicon Graphics Octane III system. See Appendix A in this guide for additional system operating specifications. Table 1-2. Silicon Graphics Octane III Physical Specifications Feature Dimensions: height width depth Specification 26.5 in. (67.3 cm) 12.5 (31.7 cm) 28.2 (71.6 cm) Weight (maximum configuration) Approximately 178 lbs. (81 kg) (max shipping 224 lbs.
Upgrading or Replacing Components • PCIe cards • Disk drives • System fans • Power supplies Service personnel should use the procedures in the guide to upgrade system components or replace failing components. Warning: If a lithium battery is installed in your system as a soldered part, only SGI qualified service personnel should replace this lithium battery.
1: System Quick Start and Overview Front view Rear view GIGE-0 GIGE-1 GIGE-2 GIGE-3 SW-CONSOLE USB-0 USB-1 VGA USB-2 Fans Figure 1-5 USB-3 DVI-0 SERIAL DVI-1 Fans Octane III Fan Locations Example Nodeboard Features At the heart of each Octane III compute tray lies a multi-processor based node board. The cluster’s serverboards are based on Intel chipsets at the time this document was published.
Upgrading or Replacing Components The following subsections describe some example features of the serverboards. See Figure 1-6 on page 15 for an example functional block diagram of a Intel based system serverboard. Processors The Octane III deskside system can support a minimum of 2 and a maximum of up to 20 processors (120 cores). The exact type of processors provided with your system depends on the specific configuration you ordered.
1: System Quick Start and Overview PCI Expansion Slots PCI expansion is available as part of specific tray configurations. Each HPC serverboard within the system has one PCI Express 2.0 x16 slot. Note that the optional graphics card is the only PCIe option with provision for external connection. Optional Ethernet Ports An optional network controller (Ethernet switch) is integrated into some models to support additional Gigabit LAN ports with (100/1000Base-T/1000BaseTX, RJ45 output).
Upgrading or Replacing Components B A B B B DDR3 DIMM DDR3 DIMM 1 2 3 B 1 2 1 CPU#1 2 CPU#2 3 3 Port1 QSFP MT25408 Connect-X IB PCI-E Gen2/DDR or QDR Port0 Ports 2,1 A 1 2 3 Kawela Ports 3,4 Ports 5,6 PCI-E x16 DDR3 DIMM A A A B DDR3 DIMM A Intel 5520 IOH36D RJ45 RJ45 Ports 7,8,9,10 ESI CLINK SST25 VF016 SPI PE5 PE4-1 DMI CLINK ICH10R SATA DDR II PCI WBD BMC/VGA VGA LPC SATA #1 SATA #2 SATA #3 SATA #4 SATA #5 SATA #6 LPCIO W83527 ACPI KBC RTL8201N PHY Dedicate LA
Chapter 2 2. System Operation and Troubleshooting The first half of this chapter describes the basic operating steps needed to keep your SGI Octane III system up and running properly. The second half of this chapter provides you with some basic troubleshooting advice. Use these sections to eliminate simple problems or obtain information that may be needed by your service provider.
2: System Operation and Troubleshooting Server Precautions If you have not already done so, review the electrical and general safety precautions in Chapter 1. Each system power supply should be connected to a high-quality power source that offers protection from electrical noise and power surges. For extra protection, use a regulating uninterruptible power supply (UPS) to protect the system’s power supplies from power surges, voltage spikes and to keep your system operating in case of a power failure.
Troubleshooting the System Table 2-1 Troubleshooting Chart (continued) Problem Description Recommended Action The Fault LED of a populated PCI slot is illuminated (on). Consult the SGI Octane III Cluster Server Component Replacement Guide and reseat the PCI card. If the fault LED remains on, replace the PCI card. The fault LED of a hard disk drive is on. Consult the SGI Octane III Cluster Server Component Replacement Guide and replace the disk drive.
2: System Operation and Troubleshooting System Fan Failure If one of the 12 fans that cool the main enclosure fails, you should be able to visually identify the problem unit by looking through the perforated sheetmetal, see Figure 2-2. The system can run with a single fan failure, but it should be replaced as soon as possible. The SGI Octane III Cluster Server Component Replacement Guide has information for service personnel to replace a fan.
Chapter 3 3. System I/O Interfaces Overview This chapter provides a technical overview of the standard and optional interfaces available on your SGI Octane III system. The major I/O connectors of the system are described and illustrated. Rear I/O Connectors As described in Chapter 1, all the system’s I/O interfaces are situated on the upper rear portion of the Octane III enclosure.
3: System I/O Interfaces Overview USB Type A Connector Figure 3-2 shows the pinouts for the USB type A connector that is used for USB ports 0 through 3 on the rear panel of the Octane III system. The USB ports are used for connecting keboard, mouse and other USB compatible devices to your system. 1 2 Figure 3-2 3 4 Pin Number Locations for USB Type A Connector Table 3-1 lists the pin assignments for each connector on the USB interface.
Rear I/O Connectors VGA Video Interface Every Octane III cluster server comes with a single 15-pin VGA video connector, see Figure 3-3 for an example and Table 3-2. The VGA interface is connected directly to the system head node.
3: System I/O Interfaces Overview RJ-45 (Ethernet) Connectors Figure 3-4 shows an example of the RJ-45 connector used on the Octane III with its pin assignments called out. Pin 4 Pin 3 Pin 6 Pin 2 Pin 7 Pin 1 Figure 3-4 Pin 5 Pin 8 RJ-45 Ethernet Connector Pin Assignments Table 3-3 shows the pin assignments for the RJ-45 connector used on the I/O panel of the Octane III system.
Rear I/O Connectors DB9 Serial Connector Figure 3-5 shows the DB9 connector pin assignments. This connector may be used as a “dumb” RS-232 style console port for connecting to the head node compute tray or (optionally) to interface with the internal Ethernet switch.
3: System I/O Interfaces Overview DVI-I Video/Monitor Port Figure 3-6 shows the optional DVI-I video port. C4 C3 24 17 9 C5 8 1 C2 C1 Figure 3-6 Optional DVI-I Port and Pinouts Table 3-4 shows the port pinout assignments for optional DVI-I port(s).
Chapter 4 4. Octane III Workstation BIOS Information This chapter describes the functions and features of the AMI BIOS Setup Utility for the SGI workstation version of the Octane III system. The AMI ROM BIOS is stored in a Flash EEPROM and can be updated as needed; check with your SGI sales or service representative for information on updates. This chapter covers basic navigation of the AMI BIOS Setup Utility screens. Important: This BIOS information is applicable to Intel Xeon based systems only.
4: Octane III Workstation BIOS Information Note: Options printed in Bold are default settings. How To Change the Configuration Data The configuration data that determines the system parameters may be changed by entering the AMI BIOS Setup utility. This Setup utility can be accessed by pressing at the appropriate time during system boot. Starting the Setup Utility Normally, the only visible Power-On Self-Test (POST) routine is the memory test.
How To Change the Configuration Data Use this option to change the system time and date. Highlight System Time or System Date using the arrow keys. Enter new values through the keyboard. Press the key or the arrow keys to move between fields. The date must be entered in Day MM/DD/YY format. The time is entered in HH:MM:SS format. (Note that the time is in the 24-hour format. For example, 5:30 P.M. appears as 17:30:00.) BIOS Build Version: This item displays the BIOS revision used in your system.
4: Octane III Workstation BIOS Information System Memory: This displays the size of memory available in the system: Size: This item displays the memory size detected by the BIOS. Advanced Setup Configurations Use the arrow keys to select Boot Setup and hit to access the submenu items: BOOT Features Quick Boot If enabled, this option will skip certain tests during POST to reduce the time needed for system boot. The options are Enabled and Disabled.
How To Change the Configuration Data This feature displays "Press DEL to run Setup" during POST. The options are Enabled and Disabled. Interrupt 19 Capture Interrupt 19 is the software interrupt that handles the boot disk function. When this item is set to Enabled, the ROM BIOS of the host adaptors will "capture" Interrupt 19 at boot and allow the drives that are attached to these host adaptors to function as bootable disks.
4: Octane III Workstation BIOS Information for the CPU installed in your motherboard is [18]. Press "+" or "-" on your keyboard to change this value. C1E Support Select Enabled to use the feature of Enhanced Halt State. C1E significantly reduces the CPU's power consumption by reducing the CPU's clock cycle and voltage during a "Halt State." The options are Disabled and Enabled.
How To Change the Configuration Data Set to Enabled to use the Simultaneous Multi-Threading Technology, which will result in increased CPU performance. The options are Disabled and Enabled. Active Processor Cores Set to Enabled to use a processor's Second Core and beyond. (Please refer to Intel's web site for more information.) The options are All, 1 and 2.
4: Octane III Workstation BIOS Information Select Enable to use the feature of Clock Spectrum, which will allow the BIOS to monitor and attempt to reduce the level of Electromagnetic Interference caused by the components whenever needed. The options are Disabled and Enabled. Advanced Chipset Control The items included in the Advanced Settings submenu are listed below: CPU Bridge ConfigurationQPI Links Speed This feature selects QPI's data transfer speed. The options are Slow-mode, and Full Speed.
How To Change the Configuration Data A memory error-correction scheme where the Processor writes corrected data back into the memory block from where it was read by the Processor. The options are Enabled and Disabled. Patrol Scrubbing A memory error-correction scheme that works in the background looking for and correcting resident errors. The options are Enabled and Disabled.
4: Octane III Workstation BIOS Information This feature defines how many meters above or below sea level the system is located. The options are Sea Level or Below, 1~300, 301~600, 601~900, 901~1200, 1201~1500, 1501~1800, 1801~2100, 2101~2400, 2401~2700, 2701~3000. DIMM Pitch This is the physical space between each DIMM module. Each step is in 1/1000 of an inch. The default is [400]. Press "+" or "-" on your keyboard to change this value.
How To Change the Configuration Data SouthBridge Configuration This feature allows the user to configure the settings for the Intel ICH South Bridge chipset. USB Functions This feature allows the user to decide the number of on-board USB ports to be enabled. The Options are: Disabled, 2 USB ports, 4 USB ports, 6 USB ports, 8 Ports, 10 Ports and 12 USB ports. Legacy USB Support Select Enabled to use Legacy USB devices.
4: Octane III Workstation BIOS Information This feature allows the user to select the drive type for SATA#1. The options are IDE, RAID and AHCI. • SATA#2 Configuration Selecting Enhanced will set SATA#2 to native SATA mode. The options are Disabled, and Enhanced. Primary IDE Master/Slave, Secondary IDE Master/Slave, Third IDE Master, and Fourth IDE Master These settings allow the user to set the parameters of Primary IDE Master/Slave, Secondary IDE Master/Slave, Third and Fourth IDE Master slots.
How To Change the Configuration Data Select Auto to allow the AMI BIOS to automatically detect the PIO mode. Use this value if the IDE disk drive support cannot be determined. Select 0 to allow the AMI BIOS to use PIO mode 0. It has a data transfer rate of 3.3 MBs. Select 1 to allow the AMI BIOS to use PIO mode 1. It has a data transfer rate of 5.2 MBs. Select 2 to allow the AMI BIOS to use PIO mode 2. It has a data transfer rate of 8.3 MBs. Select 3 to allow the AMI BIOS to use PIO mode 3.
4: Octane III Workstation BIOS Information Select UDMA1 to allow the BIOS to use Ultra DMA mode 1. It has a data transfer rate of 25 MBs. Select UDMA2 to allow the BIOS to use Ultra DMA mode 2. It has a data transfer rate of 33.3 MBs. Select UDMA3 to allow the BIOS to use Ultra DMA mode 3. It has a data transfer rate of 66.6 MBs. Select UDMA4 to allow the BIOS to use Ultra DMA mode 4. It has a data transfer rate of 100 MBs. The options are Auto, SWDMAn, MWDMAn, and UDMAn. S.M.A.R.T.
How To Change the Configuration Data Select No to allow the AMI BIOS to configure all devices in the system. PCI Latency Timer This feature sets the latency Timer of each PCI device installed on a PCI bus. Select 64 to set the PCI latency to 64 PCI clock cycles. The options are 32, 64, 96, 128, 160, 192, 224 and 248. PCI IDE BusMaster When enabled, the BIOS uses PCI bus mastering for reading/writing to IDE drives. The options are Disabled and Enabled.
4: Octane III Workstation BIOS Information • Flow Control This feature allows the user to set the flow control for Console Redirection. The options are None, Hardware, and Software. Redirection After BIOS POST Select Disabled to turn off Console Redirection after Power-On Self-Test (POST). Select Always to keep Console Redirection active all the time after POST. Note: This setting may not be supported by some operating systems.
How To Change the Configuration Data Warning: Any temperature that exceeds the CPU threshold temperature predefined by the CPU manufacturer may result in CPU overheat or system instability. When the CPU temperature reaches this predefined threshold, the CPU and system cooling fans will run at full speed.
4: Octane III Workstation BIOS Information (i.e., 25o C). The CPU Temperature feature will display the CPU temperature status as detected by the BIOS: Low – This level is considered as the ‘normal’ operating state. The CPU temperature is well below the CPU ‘Temperature Tolerance’. The motherboard fans and CPU will run normally as configured in the BIOS (Fan Speed Control). User intervention: No action required. Medium – The processor is running warmer.
How To Change the Configuration Data run at the full speed (12V). The Options are: 1. Disabled (Full Speed), 2. Server Mode, 3. Workstation Mode. Fan1 ~ Fan 4 Reading This feature displays the fan speed readings from fan interfaces Fan1 through Fan5. CPU1 Vcore, CPU2 Vcore, +5Vin, +12Vcc (V), VPI DIMM, VP2 DIMM, 3.3Vcc (V), and Battery Voltage ACPI Configuration Use this feature to configure Advanced Con.guration and Power Interface (ACPI) power management settings for your system.
4: Octane III Workstation BIOS Information Performance Event Timer is used to replace the 8254 Programmable Interval Timer. The options are Enabled and Disabled. IPMI Configuration Intelligent Platform Management Interface (IPMI) is a set of common interfaces that IT administrators can use to monitor system health and to manage the system as a whole. For more information on the IPMI specifications, please visit Intel's website at www.intel.com.
How To Change the Configuration Data Clear BMC System Event Log This feature is used to clear the BMC System Event Log. Caution: Any cleared information is unrecoverable. Make absolutely sure that you no longer need any data stored in the log before clearing the BMC Event Log. Set LAN Configuration Set this feature to configure the IPMI LAN adapter with a network address. Channel Number - Enter the channel number for the SET LAN Configuration command. This is initially set to [1].
4: Octane III Workstation BIOS Information Parameter Selector Use this feature to select the parameter of your Mac Address configuration. Mac Address The BIOS will automatically enter the Mac address of this machine; however it may be over-ridden. Mac addresses are 6 two-digit hexadecimal numbers (Base 16, 0 ~ 9, A, B, C, D, E, F) separated by dots. (i.e., 00.30.48.D0.D4.60). Current Mac Address in BMC This item displays the current Mac address used for your IPMI connection.
How To Change the Configuration Data The options are Enabled and Disabled.PEF Alert Startup Delay -This sets the pre-determined time to delay PEF alerts after system power-ups and resets. Refer to Table 24.6 of the IPMI 1.5 Specification for more information at www.intel.com. The options are: No Delay, 30 sec, 60 sec, 1.5 min, 2.0 min. Startup Delay - This feature enables or disables startup delay. The options are Enabled and Disabled.
4: Octane III Workstation BIOS Information Security Settings The AMI BIOS provides a Supervisor and a User password. If you use both passwords, the Supervisor password must be set first. Supervisor Password This item indicates if a supervisor password has been entered for the system. Clear means such a password has not been used and Set means a supervisor password has been entered for the system. User Password: This item indicates if a user password has been entered for the system.
How To Change the Configuration Data Boot Sector Virus Protection When Enabled, the AMI BOIS displays a warning when any program (or virus) issues a Disk Format command or attempts to write to the boot sector of the hard disk drive. The options are Enabled and Disabled. Boot Configuration Use this feature to configure boot settings. Boot Device Priority This feature allows the user to specify the sequence of priority for the Boot Device.
4: Octane III Workstation BIOS Information Exit Options Select the Exit tab from the AMI BIOS Setup Utility screen to enter the Exit BIOS Setup screen. Save Changes and Exit When you have completed the system con.guration changes, select this option to leave the BIOS Setup Utility and reboot the computer, so the new system configuration parameters can take effect. Select Save Changes and Exit from the Exit menu and press .
BIOS Error Beep Codes Non-fatal errors are those which, in most cases, allow the system to continue the boot-up process. The error messages normally appear on the screen. Fatal errors are those which will not allow the system to continue the boot-up procedure. If a fatal error occurs, you should consult with your system manufacturer for possible repairs. The following list of error codes my be helpful in diagnosing certain system problems.
Appendix A A. Technical Specifications This appendix contains technical specification information about your system. Server Specifications and Features Table A-1 shows the physical specifications of the SGI Octane III deskside system. Table A-1 SGI Octane III Enclosure Physical Specifications System Features Specification Height 28.2 in. (71.6 cm) Width 12.5 in. (31.75 cm) Depth 26.5 in. (67.3 cm) Weight (full) maximum 178 lbs.
A: Technical Specifications Environmental Specifications Table A-2 lists the environmental specifications of the system. Table A-2 Environmental Specifications Feature Specification Temperature tolerance (operating) +5 °C (41 °F) to +35 °C (95 °F) (up to 1500 m / 5000 ft.) +5 °C (41 °F) to +30 °C (86 °F) (1500 m to 3000 m /5000 ft. to 10,000 ft.