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

Server data recovery refers to the professional technology to repair and extract the data originally stored in the server and storage equipment, which is lost due to misoperation, hardware failure, malicious attack and other reasons. Before introducing the server data recovery, we first need to understand the data structure and file storage principle of the server. Today, through the data of Huawei S5300 server, the editor introduces the data storage structure and data recovery principle of this type of server.

[server data recovery failure description]

The failure server model introduced today is Huawei s5300 server. There are 24 hard drives in the server that need data recovery to form a set of raid5 disks, one of which is a hot spare. During the normal operation of the server, one hard disk activates the hot spare offline for data synchronization. In the process of data synchronization, another hard disk in the server is offline due to unknown failure, which leads to the collapse of the upper application of the server and the loss of data in the server.

[server data recovery detection]

After receiving all the hard drives in the customer server, the hard drives are first tested one by one, and it is found that there is no physical fault except one hard disk. Considering the security of the customer's original data, the server data recovery engineer needs to mirror all the hard drives.

[server data recovery process]

Analyze the cause of server failure

Because the controller of Huawei S5300 has a strict policy of checking disks, hard disks with unstable performance without physical failure will be kicked out of the raid array as bad disks. The server data recovery engineer detected a physical failure in one of the hard drives, so one of the two disconnected disks of the server was offline due to read and write instability.

Analyze the raid structure in the server

For the case of server hard disk offline data recovery, in order to restore lun, we first need to analyze the underlying raid structure information, and reorganize raid according to the analyzed raid array information. The server data recovery engineer analyzes each hard disk and obtains the basic information of the raid array, such as stripe size, data direction, hard disk order, hot spare, database distribution and so on.

Check and troubleshoot the synchronized hard disk in the server

According to the server array information that has been analyzed, the data recovery engineer uses the self-developed raid virtual program to reorganize the original array, but in the process of reorganization, it is found that the data in a hard disk has been destroyed synchronously. We need to eliminate the damaged hard disk in the data recovery process, and the data recovery engineer compares the underlying data structure of all the hard drives. It is found that the data of one hard disk on the same stripe is obviously different from that of the other hard drives. After using the raid check program to check the stripe of the hard disk, it is confirmed that the hard disk has been destroyed synchronously. After reorganizing the raid array, the server data recovery engineer analyzes the distribution of lun in the raid group and the data block map. As long as the map can be extracted completely, the lun data can be parsed and recovered with the help of the data recovery program.

Extract the database files in the server and repair the database

The server data recovery engineer independently wrote a file system parsing program to parse the file system in the virtual array, exported the database file and handed it over to the database data recovery engineer for verification and repair.

The database engineer validates the database files and finds that some database files and log files are abnormal. There are a large number of bad blocks in the tablespace, all the control files are destroyed, and undotbs02 is lost, so the database data recovery engineer has to repair the database files. The repair process is shown in the following figure:

[verification of server data recovery results]

After the database data recovery engineer repaired and verified the data, the server data recovery engineer successfully restored the database in the server. The server data recovery engineer imported the repaired database data into the data recovery server to verify that all the data was normal. Contact the customer for on-site data verification, and the data recovery server recovered 100% this time.

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.