Upgrading and Repairing Servers

Although most x86 and Itanium-based servers use BIOS beep codes and error messages to indicate problems, other types of servers use different methods for solving startup and system problems. The following sections highlight the methods used on some of the most popular advanced server platforms.

IBM Light Path Diagnostics

Several of IBM's server product lines include Light Path Diagnostics. Light Path Diagnostics monitors major subsystems inside the server, including the following:

  • Disk drives

  • Memory

  • Power supplies

  • CPUs

  • Voltage regulator modules

  • PCI buses

  • Processor mismatches and failures

  • Temperature alerts

  • Non-redundancy alerts

  • System board failure

The front of each server features a signal light that illuminates if any component monitored by Light Path Diagnostics has failed. Inside the server, a Light Path Diagnostics panel indicates which subsystem has failed. When the server is disassembled, an LED indicates which component has failed. Because most servers have multiple components in a particular category, such as memory modules, the ability of Light Path Diagnostics to indicate exactly the component that has failed saves a great deal of time in bringing a failed server back into operation. On a BladeCenter server, Light Path Diagnostics also features a location LED, which can be illuminated remotely by the system administrator to indicate which server blade needs attention.

The implementation of Light Path Diagnostics varies by server model. See the documentation for the particular IBM server for details.

Hewlett-Packard's RX Series Integrity Server LEDs

Hewlett-Packard's RX series of Integrity servers feature LEDs at various points throughout the system that can be used to track down problems. The front panel LEDs indicate whether a management processor is installed, how the system is being managed, power status, and system status. If a fault is detected, use the LED arrays around the system to determine the problem. LEDs provide status for the following:

  • System (bulk) power supplies

  • PCI-X power supply

  • System and PCI I/O fans and cooling status

  • Internal cell board power

  • PCI-X card slots

  • Core I/O power

  • MP I/O power

Configuring and Troubleshooting Sun SPARC Systems with EEPROM

Sun SPARC servers use EEPROM chips to store system configuration information. If this information is corrupted or missing, the system is unable to start. To view or change the current EEPROM settings, use the eeprom command from the OK prompt. You must be a superuser to run eeprom. Note that the eeprom command also works with x86-based Sun servers running Solaris; these systems use a file in the boot area to simulate EEPROM storage.

The eeprom command can be used to control and display many system features, including the following:

  • Enabling/disabling ACPI power management (x86-based SPARC servers only)

  • Configuring the console device

  • Boot configuration

  • Terminal emulation

  • System diagnostics

  • Error recovery options

  • Input device configuration

  • Hardware revision and update information

  • Network settings (MAC address, network booting, and others)

  • Path to the operating system

  • Screen configuration

  • Security settings

  • TTY settings

  • Memory testing

For examples and details, see the system administration commands section of the documentation for your server's version of Solaris.

Troubleshooting and Managing Hewlett-Packard Servers

Hewlett-Packard manufactures several server lines, including AlphaServer, Hewlett-Packard Integrity, and Hewlett-Packard 9000. Hewlett-Packard's Technical Documentation page (http://docs.hp.com/en/index.html) provides links to hardware documentation, operating system support, and hardware diagnostics for its server and other product lines.

Hewlett-Packard AlphaServers are managed with the AlphaServer Management Station (AMS) program. Some components of AMS can be run from a client workstation running Windows, and others are run directly on the AlphaServer's own console.

AMS permits you to work with server hardware and disk partitions, display events, view warnings, and perform other tasks. For more information, see the documentation supplied with your AlphaServer or available at Hewlett-Packard's website.

Hewlett-Packard Integrity and Hewlett-Packard 9000 servers use the Integrated Lights-Out (ILO) management processor to manage and troubleshoot the operations of these servers. For more information about using ILO, see the documentation provided with these servers or visit the Hewlett-Packard website.

Related

Категории