If RAID arrays collapse of what a good remedy you can think of, you learn this step to subvert the past recognition.

I. Overview of the fault memory
1, the fault environmental
groups respectively by the SAS 4 raid5 block 600G capacity hard disk arrays, and the array into two groups of LUNs, LVM configuration composition, and formatted as EXT3 file system.
2, fault analysis
an offline unexpectedly hard disk, the hot spare line, off-line to replace the hard disk. However, in the hot spare on-line process, and a hard off-line, resulting in hot spare fails synchronization, two raid array group collapse, the LVM configuration is not complete, the file system can not work properly. Two hard drives for off-line testing and found that the first off-line hard disk is not recognized, it concluded that a hardware failure, the need for opening a repair operation, a second hard drive can be identified.
Second, the solution outlined
based on the results of the pre-failure analysis, summed up the following solutions:
1, to repair the failed hard disk, use the MRT professional data recovery software for hard disk failure backup.
2, the use of professional data recovery software winhex to raid the remaining members of another set of disks and raid all members of the overall disk backup.
3, each disk data analysis, restructuring RAID array analysis.
4, analysis of recombinant finished array, find information LVM, the LVM volume recombination.
5, EXT3 file systems on LVM volumes restructuring parse, restore and export all the data.
Third, implement the solution
1, fault repair disk
failure hard disk repair operation opening. After opening it found that hard disk platters badly worn beyond repair, only to lack of disk array processing.
2, hard disk backup
using specialized data recovery software winhex raid on the remaining members of an array of disk failure backup overall, and for all members of another set of disks raid array is good overall backup, the backup as follows:


3, a recombinant raid array
Careful analysis of the underlying hard disk data, by analyzing EXT3 file system structure, the two sequences were analyzed disk raid array, stripe size, configuration information such as direction check, using professional data recovery software winhex two sets of recombinant raid arrays. After analysis, the two groups are raid array block size of 64K, the direction for making the synchronization check, be noted that the lack of disk failure processing raid restructuring.


4, recombinant LVM configuration
after the two recombinant raid array, the underlying data sets were analyzed in raid, find LVM configuration information, analyzes the structure of the LVM, as the raid two LUN PV (LVM physical volume) is derived and then use a professional data recovery software UFS Explorer two PV restructuring to regenerate the LVM logical volume.


5, restore data
after reorganization LVM, the pair of LV (logical volume) to parse EXT3 file system, restore and export all the data. The following is a reply out of the data:

Fourth, data validation
on the recovered data, choose some compressed files, verify and found that some files are corrupted, the results of the analytical results and the recovery contrast, found that some files are damaged and can not be recovered. After analysis, it concluded that there is file corruption and bad sectors of the hard disk in the raid two related. The following two groups of bad sectors of the hard disk where the raid:
Raid. 1: # 2 67 Bad sectors Source encountered.
. 4 Source # Bad sectors 13 is encountered.
Raid 2:. 2 Bad # 37 [Source sectors encountered
V. Conclusions recovery
due to failure hard disk corruption serious hardware can not be repaired, and the presence of the hard disk bad sectors, leading to a possible defect in the structure of the raid, part of the file is damaged, but most of the files validated recovery is successful, only a small part of the file is missing or damaged, the successful completion of this data recovery .