I have 1 cluster with serveral p4500.
There is HW Raid5 and the Raid5 is split on: first 6 is one raid 5, and the other 6 is one raid 5.
There were two disk that came up with a fault.
I replaced both disk in same Raid5.
Now the storagesystem says that Volumes is unaccessable and the Raid5 of the P4500 is down.
How can I get the lefthand up and working again? Ive tried to put the old disk in the P4500 again, but this does not help.
I also tryed to restart the disk shelf as well.
Do you have any idea? Or something to get it to work again?
Best Answer
P4500s in RAID 5 actually do a RAID 50 across the 12 physical disks in the shelf. It's a 6 to 6 split that is then striped across. They are not independent RAIDs. This means that if you lost one array, all of the data on that shelf is gone.
The good news is that you say this is a member of a cluster. If it's a mirrored cluster like it should be (Network RAID 1) then popping the new disks in should cause there to be a rebuild from the other nodes in the cluster. If it doesn't happen automatically, consult your manual. I've never has a whole array fail in a P4500. A full rebuild from another node can take a long long time. In the management console, you can adjust the max throughput for this operation. The default setting is 16MBps. Keep in mind that adjusting this can impact production.
If you set this up as Network RAID 0, then all of the data on all cluster nodes is lost. Start pulling your tape backups, you have a lot of work ahead of you.