After SAS builds a raid, the system fails to start and the recovery succeeds

Case:
A old IBM tower PC Server uses four 146G SAS hard disks to create RAID, RAID cards create two RAID levels in the disk, the front is RAID1, which is used to install the system. The RAID5 is used as another other.Logic partitions are used, and users only say that the system can only be caused by the system.

Solution:
1. Remove the hard disk and make a mirror of the hard disk.2. During the mirror process, there are two disks that have different degrees of bad sectors. For hard disks with bad sectors, we switched to PC3000 for UDMA for mirror operations.3. Analyze the underlying structure, and find that two disks with bad sectors have been created by the RAID card as RAID1 since the 0 sector. That is to say, the server uses two plates of the four plates to create the RAID1 to install the system.The starting sector structure of the remaining two plates is RAID5. The two disks are created with RAID5 as logical partitions with the remaining sectors behind the two disks containing RAID1.4. Establish RAID, use the software we developed independently, and position the same RAID -level sector address in the hard disk to virtually build RAID and finally restore user data.