Shulou(Shulou.com)06/03 Report--

[background introduction to server data recovery]

A HP server, attached to a HP MSA50 disk array, with 5 1TB hard drives, the original structure is RAID5.

After using it for a period of time, one of the hard drives is offline, and the data is not wrong because RAID5 supports the redundant protection of a hard disk error. Then after running for a short time, the server failed, so the maintenance personnel did not fully understand the situation, and recreated a new set of RAID5 for the remaining four hard drives and completed them synchronously, resulting in the loss of all the original data.

Several data recovery companies have processed it and cannot recover it.

[data recovery failure analysis]

When HP SMART ARRAY creates a new set of RAID5, it rebuilds all block checks by default, that is, in any band that makes up the RAID5, there is always a check block whose data is generated when it is created, which is corrupted relative to the original useful data. After analysis, the generated 4-disk RAID5 is organized in the way of double-cycle, 64K block size and 16 stripe change check, that is to say, about 1m of data every 3m of the 4 disk members is wrong.

At the same time, the analysis shows that the composition structure of the original 5-disk RAID5 is double cycle, 128K block size and 16 times strip change check.

In order to recover data, we must first repair the early offline hard disk, and the recovery rate can only be restored to the combination of the early offline disk and the new disk, depending on whether there are many data changes after the early disconnect.

The solution is that, through the analysis of the difference of the composition structure before and after the two times, the former offline disk is used to make up the check information destroyed during the reconstruction of the RAID, and then the RAID is virtually reorganized, the file system is explained, and the file is exported.

[record of data recovery process]

1. Backup source media

2. According to the data traces before and after destruction, the RAID structure before and after destruction is obtained.

3. Analyze the difference, write the check correction program, and virtual reorganize according to the previous RAID structure to generate the reorganized image file.

4. Fix the reorganized image file system error (because the data changes very little, the help error is very few).

5. Some partitions export data, and some partitions are completely mirrored to the new space without error.

6. Testing and acceptance

[conclusion of data recovery]

After 24 hours, the data was recovered successfully, with a success rate of nearly 100%. Users are very satisfied with the results of data recovery.

[data recovery performance pledge]

1. Free testing

Sign a confidentiality agreement with the customer (non-customized free of charge), strictly keep the customer's data confidential, recover the data unsuccessfully, and provide services free of charge by professional engineers, while other engineers review, consult the data recovery plan and quote before the process data recovery. After customer confirmation, the engineer will start the data recovery process without any write operation to the customer's original disk to ensure that the data of the original disk is complete.

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.