Server data recovery environment:
a DELL EqualLogic PS series storage. There is a RAID5 set of 16 SAS hard drives in the storage. The upper layer is the VMFS file system, which stores virtual machine files. The upper storage layer is divided into 4 volumes.
Server failure & detection:
2 hard disk indicators on the storage are yellow. The disk has failed and the storage is unavailable. The storage device has expired.
Hardware engineers performed hardware fault detection on 16 hard disks in the faulty storage and found that 2 of the disks had bad sectors, and the SMART error redundancy level had exceeded the threshold.
Server data recovery process:
1. Perform sector-level full-disk mirroring on 14 normal disks in read-only mode. For the two disks with bad sectors, use professional tools to process them and generate image files. After the mirroring is completed, all disks will be restored to the original storage as they are. Subsequent data analysis and data recovery operations will be based on the mirror files to avoid secondary damage to the original disk data. 
2. View the underlying data based on the image file and collect the log information stored in the fault. Analyze the collected log information, find out the offline time of the two failed disks, determine the disk with newer data, and use the hard disk with newer data to restore the data.
3. Analyze all disk underlying data based on the image file, obtain RAID5 structure-related information, and virtually reconstruct the original RAID5 based on the obtained RAID-related information.
4. Beiya Qian's data recovery engineers extracted 4 LUNs from the virtual RAID5 through bitmap information.
5. Combine the four VMFS file systems into spanned volumes according to the underlying structure, export the data, and verify whether the virtual machine is normal. 
6. Copy the files in the volume and verify the restored virtual machine through network sharing. The virtual machine can start normally.
7. Migrate the virtual machine files to the environment prepared by the user in advance. After testing by the user, it is confirmed that the recovered data is complete and valid, and the recovery results are approved. This data recovery work is completed.