I was called into a business that had an emergency data loss situation on their RAID 5 server. The server (an HP Proliant) could not load the operating system and the raid partition appeared as unallocated.
Data and file loss on servers is always of high priority because these computer contain most of the data used by a business, so when the server stops working, so do the people as the server contains the data that they are working on.
Now HP Proliant servers have a good reputation as computer servers: they most run a distributed file system known as raid too which is good as it distributes data across the hard drives present in the server.
In this case there were 5 disks configured in a RAID 5 system. The server was switching on OK, but the RAID partition wasn’t coming up. I inspected each of the hard drives and found that a Seagate Berracuda 2TB hard drive stopped working recently and was producing a beeping/clicking sound. This usually means that the disk has failed with a mechanical error. This problem was also stopping the machine booting.
RAID 5 servers should in theory continue working if one of their hard drives fails, but this time it was obviously not working. The disk was replaced (‘swapped out’) and the RAID information rebuilt (an automatic process). This took an hour or so as data from copied (‘rebuilt’) onto the new disks from the other 4. Once completed the RAID was rebuilt, the server was working again and the data accessible.
Learn more about my and Data Clinic’s raid and server data recovery services at http://data-recovery-tips.co.uk/raid-data-recovery/.