We recently had a raid array fail on a SQL 2008 server with 8 x 73GB drives. (Dell 2950). There were two drives that went bad (one had been bad for a few months it seems and was not caught...). To top it off log shipping was not working on this server and the last back up was 3 days old…
So I have these 8 drives on hand, which ones (if any) have specific completed DB files on them if I choose to send them out for recovery?
My programmer says HDD2 and HDD3. Is there a consistent logic used by Raid 5 or SQL to write data to a specific drive in these cases? HDD0 died first, and then HDD4 was the last drive that failed and crashed the array.
On a side note, we will no longer be using Raid 5 and moving to Raid 10 now on all new servers.
Thanks in advance.