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

[fault description]

A local court has a HP EVA8400 storage and two sets of expansion cabinets. The physical disk is composed of 12 1T FATA disks (AG691A 454414-001) and 10 300g 15K FC disks (AG690A 454411-001). The number of LUN is uncertain, the host environment is WINDOWS, and the court historical case trial materials are stored.

As this case changed hands in many ways, we can not directly know the cause of the failure.

[preliminary inspection and analysis]

1. After the initial check by telephone, it is confirmed that the data is not reused after failure. Usually according to the failure of HP-EVA, it may be inferred that the reliability of data recovery is high.

2. The EVA host and expansion cabinet shut down normally, then mark the location serial number of all the hard drives and take them out. The EVA 8400 controller is no longer powered on until the data is successfully restored.

3. After taking over the disk, connect the disk according to the following link.

4. Enter the WINDOWS environment, check the disk situation with WINHEX, and find that all disks can be identified normally.

5. Check the information of each disk and find that there is PV HEAD on 300g FC disk and no PV HEAD on 1T FATA disk. Looking at the Metadata stored on 300G disks, it is found that only one RSS group of LUN is described, which is less than 2T in size and consists of all 300G disks. On the other hand, the residual LUN information in 1T FATA disk includes at least 5 sets of information. The above information indicates that, most likely, all the VDISK partitioned within the DISK GROUP composed of 1T disks were deleted and all 1T FATA disks were UNGROUP for some reason.

6. By analyzing the Metadata retained on 1T FATA disk, we can roughly judge that the recovery rate is higher.

[recovery process]

1. To make a full mirror of all disks, see "how to make a full backup of disks" or other articles in my blog.

2. Use Frombyte recovery for HP-EVA to restore the LUN to which the 300g disk belongs.

3. Because the 1T disk has all UNGROUP, it is impossible to know the allocation of RSS and its own disk ID. Therefore, it is necessary to analyze the RSS configuration table manually. Through the comparison of META information and the verification of xor information area, the following rss group configuration table is obtained:

3-0 hd6

3-1 hd8

3-2 hd2

3-3 hd9

3-4 hd10

3-5 hd5

2-0 hd0

2-1 hd7

2-2 hd1

2-3 hd11

2-4 hd3

2-5 hd4

4. Reorganize and integrate all LUN storage allocation tables.

5. Extract all LUN according to the storage allocation table and the RSS disk allocation table. In the process of extraction, the failed XOR stripes are manually analyzed to determine the offline situation (there is no offline disk in this case), determine the best reorganization conclusion, and then recover through Frombyte recovery for HP-EVA.

[conclusion of data recovery]

100% of the data was recovered successfully.

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.