Case:
Poweredge R710. The disk system is RAID5 composed of SAS 1T*3.Linux system, after FSCK, the system cannot mount.After investigation, RAID5 is intact. After FSCK, the bitmap and descriptions in the first group group and the first block group are filled by garbage logs. The root directory in the first block is also filled by garbage logs.I can’t know all the information of the file system.
Solution:
According to the existing file system node and residual log area, the engineer restores the original Superblock; according to SuperBlock analysis, the file system is drawn to EXT3, and its original log node is 8. Based on the disk structure, the starting position of the log node (superblock of the log information and the log page of the type 0x04) obtained its size, and reverse the analysis of its INODE, which is based on the root directory node. Transferred to the root directory area, about 500 LBAs (usually near this) have been filled by garbage logs; the root directory record is restored from the log. Restore the possible inode in the first group. The system structure of the file system can already be read under Linux. Most data can be read, but most of the important data required by the user is wrong. It is suspected that the collapse in the use is caused by the characteristics of EXT3. Most unable to read nodes were filled in log garbage. Analysis in the log file, if you can retrieve, generate a good node and copy it; if the log has no reference, analyze it through the data area structure. (All the directory areas are intact in this case) The recovered data is accepted.