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

1. Server device type:

HP DL380 server

Three pieces of 300g SAS disk

Second, the description of server failure:

The failure of the hard disk paralyzes the entire RAID group, which is shown as one of the hard disk status lights is red. Since the database is stored in partition D, the backup is stored in partition E. Storage failure, resulting in the D partition can not be identified, E partition can be identified, but the copy backup file reported an error, by restarting the server, resulting in the first offline hard disk online, and synchronized for a period of time, directly forced shutdown without the completion of synchronization, and then did not do anything to the server.

Third, the operation flow of server data recovery:

1. In order to ensure the safety of the data in the existing disk, first make a read-only mirror backup of the disk, three hard disks can be read normally, and no bad paths are found. The read-only mirror backup log is shown in the following figure:

2. Analyze the backup image files in detail, reorganize the raid structure, and do XOR check, and some of the checks pass. Due to the offline hard disk online synchronization operation, will damage the data, so as shown in the following figure indicates that the data is damaged.

3. In the process of RAID analysis, try to extract data from a variety of hard drives offline, and the data extracted from each disk is the same.

4. Firstly, analyze and repair the dat file in E partition. Both backup files are found to be corrupted, as shown in the following figure:

5. Analyze the aggregated dat fragments to verify the integrity of the dat data, and the underlying structure shows damage.

6. At the same time, analyze and scan the data files of D partition. Due to storage synchronization, the data file directory is not visible.

7. Scan the free space data page of partition D, and analyze and aggregate the file fragments.

8. Verify the integrity and validity of data file fragments.

9. Extract the data from the backup file and record it into the new database.

10. Connect the database through the upper application to verify the data availability, the database file can be loaded normally, the user account in the upper application software is normal, and the normal data query can be carried out.

4. Server data recovery results

After XOR detection of the reorganized raid structure, it is found that part of the data is damaged synchronously, but some of the data is normal.

During the recovery process, 2 backup files of SealLib databases were found on disk E. However, a small part of the page structure in the backup file data is damaged.

In the results of the D-partition scan, more continuous data fragments are found in the data fragments, and the fragments are available. Through the integration and splicing of D partition fragments and E partition backup files.

Fifth, server data recovery conclusion:

Through the engineer parsing processing, the final repair of the parsed data can support the normal use of the whole application, and the upper application can query the database content normally, so that the data recovery work is finished.

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.