RAID5 array strip analysis and array restructuring process

1. Server RAID 5 disk array failure

The server model is p2000 server/RAID 5 disk array/10 disks/single disk capacity 1 TB/Hot Standby disk number 6. The fault is that two disks in the disk array are offline, the system prompts the ××fill light. The initial check result is that the serial number of the array disk cannot be read/identified by the expansion card, and the physical failure of the disk is preliminarily inferred.

2. Raid disk array data backup

My company's server data recovery engineers performed a preliminary check on the faulty device and found that the disk array was faulty in a complicated situation. They needed to use different methods for handling the normal and offline hard disks, backup images can be used to back up normal disks in the raid Array Using data recovery software. For offline hard disks with physical faults, You need to repair hardware faults Before performing backup. the commonly used disk fault repair method is power-on detection, determine whether the faulty hard disk has been started or not, and whether the head can be traced, and then clean the hda oxidation component, replace the ROM chip, replace the head, and other common hard disk data recovery methods. However, in the actual data recovery process, hardware damage to one disk is serious and cannot be repaired. You can only back up and restore data through another disk.

3. Data Recovery by reorganizing the RAID5 Array

3.1.Determine the start sector
Open all nine disks that do not include the hot spare disk by using the data recovery tool and resolve the image file as a disk. It is found that the 0 sectors of all disks in the RAID 5 array are "55aa ", "05" is displayed at 0x01c2h, indicating an extended partition. The MBR partition is incorrect. Continue to search for this array and find that the MBR partition of one disk is normal on the other disk (the value at 0x01c6 indicates that the next sector is the GPT header ), it can be determined that the disk is the first hard disk, and the sector where the GPT partition is located starts from 172032 sectors. Therefore, it is preliminarily determined that the starting sector of the Lun is 172032 sectors.
. Determine the size of the raid array stripe (strip)
Strip is the basic unit used for data processing in a raid disk array. The size of the Strip varies depending on the raid disk array, the reason for analyzing the Strip size is that each strip size in the RAID 5 disk array contains an equal verification area. Therefore, the size of the strip in the raid array is 1024 sectors.
3.3 determine the disk sequence of the disk array
As the stripe size of the array has been analyzed to be 1024, this rule is used for segmentation, so that each stripe is the same as the record size and all disks in the array are jumped to 283123. Then, by comparing and analyzing the direction of the RAID5 array, you can put the first hard disk that has been analyzed first, the actual array disk sequence is 9-3-10-4-2-1-5-8-7.

3. reorganize the raid Array

The raid array is reorganized using the data recovery tool. The attribute is RAID5/stripe size 512kb/left asynchronous. Data verification can be performed after array reorganization.

4. data verification/RAID5 data recovery successful

The data of the restructured array is successfully verified. The data is successfully restored.

RAID5 array strip analysis and array restructuring process