Case:
Customer storage model is IBM DS3400, using 12 300GB SAS disk RAID5, which has no heat.There are 3 Luns on the storage, two of which are LUN where the Oracle production library of the ERP system is located, and the other is LUN where the Oracle database backup file is located.Lun’s use environment is RHEL5.Because the two disks are offline, the RAID collapse is caused, and the data is lost.After detecting the hard disk, a large number of bad sectors were found in many hard disks.
Solution:
1. After each hard disk is treated, the mirror is fully mirroring into a safe storage.2. All original disk disks of the customer to ensure that it is not damaged by the outside world.3. Analyze the original RAID combination parameter from the disk mirror.4. Analyze the first offline disk, because the first offline disk is not involved in the storage of the data after the offline disk is offline. The data written later has nothing to do with this disk. Therefore, this disk must be selected. This disk cannot participate in data recovery.5. Virtual reorganization RAID and restore all data.The ORACLE database was later verified, but it was found that a table space file was damaged. This was caused by disk bad sectors. After the database recovered engineer was repaired, one was normal.
