Case:
The customer server is Shuguang server, with unknown models, a total of four hard disks, 146G per block, which is the SCSI interface. One day suddenly crashed. After restarting, the RAID card prompts 2 hard disk errors. The customer also has a good server of the same model. ThenAll hard disks are not successful in order to another server.The internal database is very important, and no other operation is done.
Solution:
The engineer tested each hard disk on the special work machine. It was found that there were a small amount of bad disk on the No. 2 and 3 hard disks. All hard disks were mirrors. The mirror was used for 3 hours.It’s relatively smooth.Analyze the main parameters of the original RAID5. It takes 15 minutes to directly load and successfully restores all database files. The directory is complete.100%recovery.The maximum database file, more than 30 g, can be directly attached in SQL Server, DBCC has no error.