Troubleshooting of hardware faults on old servers (1)

Hardware faults refer to various errors caused by server hardware exceptions. Because the composition of the server is complex, you must check the server carefully and carefully. The following uses an HP LH6000 as an example.

There is an HP LH6000 with M memory, using a PIIIXEON700 processor with 2 m high-speed cache. No display is displayed after the instance is started, but the system log displays a message indicating that the CPU voltage is 0 V. The system indicator lights are no longer blinking. The three lights are flashing, which is another alarm method for the server, I will explain it later ). This error is generally caused by an error in the processor Voltage Regulator Module VRM) or a CPU error or poor contact between the CPU and the CPU, but it may also be caused by an error in the CPU. This situation is complicated, you must think carefully. Because the CPU segment plays an important role in the entire server, if it fails, the server will report a fatal error and the system log will prompt a fatal error, however, the error message about CPU voltage is about 5%. We immediately switched the CPU to another CPU slot, and the fault was still the same after the boot. Therefore, in the preliminary judgment, it can be ruled out that the CPU segment is bad.

At this time, after the CPU is carefully wiped, And the CPU section is in contact with the CPU, the boot is still not displayed.

When the processor is poor, the processor voltage module VRM is faulty. Therefore, the next processor voltage module is immediately obtained from another LH6000 and installed on this server. After the server is turned on, the system logs still show that the CPU voltage is 0 V, and the system light is still flashing. At this time, the situation is obvious. Therefore, after the next CPU is immediately removed from the other LH6000, the instance will start up normally.

Clues may be confusing during server maintenance. Generally, it is impossible to accurately identify the problem at a time. This requires the relevant personnel to have confidence and patience. When an error occurs, the general process is solved through the information in the system log. If the problem is not solved, find other factors and then read the log information. In short, after a server error occurs, it must be solved step by step. There is no shortcut.