[server data recovery] a case of RAID5 data recovery stored in a brand of server

Server storage failure ：

A brand of server storage crashed due to unknown reasons , When troubleshooting, the administrator found that two hard disks in the storage were offline , Hard disk location raid5 The array cannot be used , Data loss in storage , The administrator contacted our data recovery center to recover the raid5 Data in .

Server storage data recovery process ：

1、 The data recovery engineer rushed to the scene to deal with the fault raid Physical detection of disks in , Find out what to do raid in 2 None of the offline hard disks has hardware failure .

2、 The data recovery engineer will first raid Mirror and backup all disks in , The backup files are stored on the data recovery platform of the data recovery center , Return all disks in the original storage to the administrator . All data recovery operations are carried out in the data recovery platform , To ensure that the original data is not modified and destroyed .

3、 The server data recovery engineer is right raid Image file analysis found this raid There are two hot spares in , When the hard disk is offline, only one hot spare disk is successfully activated , here raid5 The array is still in a disk shortage state , Data is not synchronized .

4、 Data recovery engineer analysis raid Hard disk distribution in 、raid Strip information 、 Disk sequence information . According to the above analysis RAID Information analysis of the data in each hard disk , The data recovery engineer found that the data of a hard disk on the same strip was obviously different from that of other hard disks , Therefore, it is preliminarily determined that this hard disk may be the first to be disconnected . Data recovery engineers use RAID The verification program verifies this strip , Finally determine the hard disk that drops the line first .

5、 The data recovery engineer obtains raid Information reorganization raid, Through reorganization raid analysis lun Allocation and data block of . Data recovery engineer exports with software lun And prompt an error when parsing the file system , After re debugging, the error still exists . After troubleshooting the file system parsing error caused by software failure , The data recovery engineer manually checks the exported files , It is found that the reason for the software parsing error is that the file system metafile is damaged . The reason for this may be when the storage is paralyzed zfs The document is in progress IO operation , As a result, the file system meta file is not updated and the meta file is damaged . Because the file system cannot be parsed by software , North Asia data recovery engineers can only manually zfs Repair the damaged metafile in the file system , And then analyze it .

6、 Use the software to parse the repaired file system again , Successfully resolve all file nodes and file directory structures , Export data .

The server stores data recovery results ：

The administrator verifies the exported data on the data recovery platform of the data recovery center , Confirm that the data recovery was successful .