The software engineer encounters hardware fault. RAID5: two hard disks offline

An IBM x3650 M2 server with six hard disks (SAS 146 GB) consisting of RAID 5 with no hot backup disk. The system runs normally without an error warning. After shutdown, the system cannot find RAID and cannot start, the hardware reports a DASD error when the two hard disks are highlighted with yellow lights. The hardware supplier's repair engineers were asked to go to the door for inspection and found that the RAID was in the offline, PD Missing, and failed status, the conclusion is that the two hard disks are damaged and cannot be repaired. We recommend that you transfer them to a third-party company for data recovery.

Since there is no experience in handling similar problems, in order to maximize the protection of data, we did not try to recover the data, and we handed the data to the data recovery company for data recovery, complete the data image to the backup hard disk (about 2 hours), sample verification data is completely restored.

With data security, the system tries to recover the server by itself. During the self-check process, the RAID self-check is invisible, and enters the web bios management interface through BIOS settings, select Disk 3 (the hard disk should be taken offline and restored to the conclusion given by the company after analysis), and change its status to "Unconfig Good ", after you save the settings and exit, restart the system (the yellow light on Disk 3 is off at this time). You still cannot see RAID self-check. However, after you enter the web bios page again, you can see that the RAID status has changed to Online and degraded, and the status of Disk 3 becomes normal (Online ).

Do not give up, start random attempts (this step should have no reference value), put the system installation disk into the optical drive, choose to start the device as the optical drive, the prompt message "press any key to start from the disc" appears, and the Windows startup interface has been displayed for a long time.

After the server is started, it returns to normal operation and provides a warning dialog box for RAID downgrade. At this time, Disk 3 is normal, and disk 4 is still on with a yellow light, then, the No. 4 disk is replaced in the heat engine state, and the server starts Rebuild automatically. Since then, the server has completely recovered to normal and the original application has been operating normally.

Summary:

1. RAID5 supports only one hard disk offline. Once 2nd hard disks are offline, they enter the offline status. replugging the hard disk will not be automatically restored. You need to manually "Force launch" the hard disk after it is offline.

2. In the event of RAID damage, if you are not familiar with data recovery, try to protect the site. You can transfer the data to the data recovery company for data recovery (paid). The probability of data recovery is still high. (Even if hardware damage occurs on the hard disk, it may be restored)

3. In this example, the 2nd hard disks are not damaged, but the offline reason is uncertain. It is strange that the server has no error prompt when the RAID is downgraded due to the damage of the 1st hard disks. (Hardware inspection is not well performed)