Data recovery from a storage system crash in a scientific research computing center

Case:
Customers In order to calculate the scientific research center, the disk cabinet storage of a scientific research project computing platform is damaged, causing the calculation data to be unable to access.The specific environment of the customer is that the host server is two Dell 2U server, stored as Dell MD3200 header, with 3 MD1200 cabinets, divided into three sets of LUN nodes for server use.The server OS is the Reahat 5 Enterprise Enterprise, which runs a set of scientific test data and uses a specific dedicated database system.

Solution:
1. Data recovery company engineer testing fault storage, which consists of 12 2TB hard disks, 20TB data space, and the file system uses XFS form.2. Detect two offline fault hard disks and repair the recent offline hard disk.3. Virtual reorganization RAID, analyze the XFS file system, and export test data.4. Customer reconstruction failure RAID, the file system uses a mature and stable EXT4 system.5. Direction data, write batch script, and re -give each user corresponding permissions.6. Test data, formally restore the production environment, and the entire data is restored.