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

The server data recovers the hardware state:

The customer's server model is Huawei 5800, in which a total of 10 hard drives constitute a raid6 disk array for internal use, and the server uses the EXT3 file system; it is divided into 2 lun; and each lun8TB size.

Failure analysis of server data recovery:

During the use of the server, the administrator found that the raid failed, so he reassigned the raid to the failed server, initialized the raid at the same time, and forcibly stopped the initialization when the initialization reached about 40%, but some of the data has caused irreversible damage. It is more difficult to recover data.

Server data recovery success rate estimate:

The reason for the loss of server data is the failure of raid, and the administrator then reassigned 9 hard drives in the raid6 array to riad5 array and initialized it for a long time, which is irreversible damage to the original data. Later, in the server data recovery operation, it is also proved that only the second LUN can recover the data using the ordinary RAID6 data recovery method, but the important data that the customer needs is concentrated in the first lun. The possibility of data recovery is extremely low, and a number of data recovery companies have been involved before being connected to the customer server, all of which failed to recover valid data successfully.

Server data recovery process:

1. Quickly analyze the RAID and disk organization of the original disk RAID6 in the server. Then analyze the organization structure of RAID and disk when the server reassigns RAID5. In the actual operation, due to the redistribution of the underlying RAID6 and RAID5 information overlap, it is very difficult to analyze and distinguish these data, and the service data recovery engineer spent a day to analyze.

3. Judge the recoverability, design the algorithm to realize the recovery program and test it. After analyzing the information of the original raid6 array and the redistributed raid5 array in the server, the engineer studies the data recovery algorithm and finds that the original data of the server can be recovered by other ways. So put into the programming and correction algorithm work, the server in the original raid6 array. The first and second LUN are mirrored to the storage of the two 7TB, respectively.

4. Restore server data.

The server data recovery engineer verified that the second LUN data is completely normal, but there is about 10MB data destruction in front of the most important first LUN. The former 10MB data is extremely important. The root directory of EXT3 and the I nodes of the first block group are all in the front 10MB. Engineers try to scan and restore with the help of several commonly used software for data recovery, but the recovery results are not satisfactory. This must be the reason why the previous data recovery companies did not succeed.

In this case, the corrupted EXT3 file system has to be repaired. First of all, make a Mini Program to look up the orphan directory of the EXT3 file system, and find 3 subdirectories under this directory. Rebuilding the root directory and I node, opening it with the file system parser has been completely normal, but in order to ensure some permissions and attributes of the original data, LINUX can be mounted normally in LINUX simple repair, and then on the partition of the single disk of EXT3 where LINUX copies and formats the files with the cp command. In this way, when customers use the data, they no longer need any other settings, and after direct cp, the file directory structure and properties are exactly the same. The data was recovered successfully, and the available data is 100%.

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.