Shulou(Shulou.com)06/02 Report--

[introduction to Raid5 Array data recovery case]

Record a successful case of raid5 disk array data recovery today. First of all, a brief introduction to the basic configuration of the server that requires data recovery:

The customer's server device is EMC storage, raid5 disk array, which is composed of multiple stat hard drives and contains two hot spares. When the server crashed, two of the underlying hard drives failed and one of the hot spares was activated.

[server data recovery routine detection]

The data recovery engineer of the data recovery center goes to the customer site to detect the failure of the server equipment. The server data recovery engineer simply checked the client and server equipment to confirm that the raid5 array was down; the upper lun could not be used properly, and one of the 2 hot spares had been started.

The data recovery engineer conducted a physical inspection of the disconnected hard disk, and no physical damage such as bad track or head failure was detected in both hard drives. Subsequent data recovery operations can be performed without the need for physical repair.

[raid5 data backup]

Before the data recovery operation, all the original data on the server device needs to be mirrored and backed up. In this data recovery case, the server data recovery engineer connects all the hard drives to the data recovery dedicated storage pool in the data recovery center and mirrors all the hard drives at the sector level.

[analyze the server raid group structure]

The server data recovery operation is usually based on the recovery raid group, so this data recovery operation also needs to analyze the basic information of the raid array in order to reorganize the raid structure.

After the analysis of each hard disk by the data recovery engineer, it is found that there is no data in the two hot spares in the customer's original server, that is, even the activated hot spares are also not synchronized to any data. In fact, the two hot spares did not have any effect. If you want to recover the data, you have to continue to restore the other basic data of the original raid5 array and reorganize raid5.

[analyze the disconnect disk information of RAID group]

According to the above data recovery idea, the server data recovery engineer uses the self-developed data recovery tool to analyze the basic information of the group of raid5 array, and virtual reassembles the raid5 disk array. Under the normal process of raid data recovery, we will remove the earliest disconnected hard disks from the array where multiple hard drives are offline. Usually, the method of data analysis is to compare whether the data of each hard disk is consistent on the same stripe, strip check after removing the obviously different hard disk, until the best state of data recovery is found.

[analyze LUN information in RAID group]

After successfully reorganizing the raid5 array, the engineer began to analyze the lun information. In this data recovery case, there is only one lun on the upper layer of the client's server, and the engineer only needs to analyze the unique lun information, and then use the raid data recovery program to parse and export the map of the lun data.

[ZFS file system parsing and repairing]

Use the North data recovery tool to parse and restore the upper file system. The upper layer of the customer uses the zfs file system, and the server data recovery engineer found that some of the file system metafiles reported errors when parsing the file system, so the data recovery engineer debugged the existing data recovery tools with debug to make the program adapt to the actual situation of this data recovery.

After debugging, the reason for the error in the parsing of the zfs file system is that some metafiles in the file system are damaged due to the sudden paralysis of the client's server, resulting in the failure of the data recovery tool to parse normally. Server data recovery engineers use manual repair and other methods for damaged meta files to ensure that the zfs file system can be parsed normally.

[export all data in raid5 array]

Through the complete analysis of the file system, the server data recovery engineer finally exports the data in the customer raid array, and the customer engineer cooperates to build a data verification environment, and all the recovered data is verified. After verification, all the data in the customer's original server are successfully restored, and this data recovery is successful.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.