RAID 5 offline hard drive data recovery-server disk array data recovery method, RAID 5 disk array

RAID 5 offline hard drive data recovery-server disk array data recovery method, RAID 5 disk array

[Disk array fault occurrence process description]The customer's DELL MD1000 server has 15 built-in 1 TB hard disks built into a RAID 5 disk array. The server normally has a hard disk offline, and the Administrator synchronizes the disk array, when the disk is synchronized, an alarm is triggered on the second disk indicator. Disk Array synchronization fails, and the server crashes and cannot work properly.

[Initial inspection result of data recovery Company]The data recovery engineer checks the status of the physical disk on the server. The first offline hard disk access speed in the disk array is extremely slow. The second offline disk has a small number of bad sectors, and other disks have no obvious physical problems. The logic of the disk array is as follows: the entire disk array contains only one volume group, which occupies all the space of the array; the volume group has only one XFS bare partition with the starting position of 0 sectors.

[Disk array data recovery ideas Analysis]The RAID5 array architecture is designed to support error redundancy for only one disk. When the second disk is offline, the array will not work properly, therefore, the crash of the entire server is mainly caused by the offline operation of the second disk. Therefore, the processing of the second disk is the key to server data recovery.

 

[Server disk array data recovery process]

1. perform an exception or test on 15 hard disks on the server. All disks pass the test without any obvious error.
2. Back up the key areas of all hard disks, separately back up the second offline hard disk, and slightly over the bad sector during the backup process.
3. Calculate the data at the location of the damaged sector of the second hard disk and write it into the image file.
4. Analyze the original RAID structure during the backup process to build a virtual RAID environment.
5. Verify that the RAID structure is correct.
6. Back up the image of the second disk to the new disk, force it to go online, replace the first disk, and synchronize it. (Note: This process is dangerous and you must be cautious when performing this operation !)

7. Copy data.

[Server data recovery conclusion]Because the exception or test is completely passed, it indicates that no new data is written or the structure is changed after the storage fails. Therefore, you can calculate the data of the corresponding location of the bad disk based on several other hard disks. After the restoration is completed, the directory structure is complete and all important documents are intact. FSCK has no error message. The customer accepts the recovered data and the data is restored successfully.