I have Bl465 C g1 server with only two hard drive slots which are currently occupied with 72GB disks (physical) which are at raid 1 level. This contians the OS partition aswell now I need to replace t
I have Bl465 C g1 server with only two hard drive slots which are currently occupied with 72GB disks (physical) which are at raid 1 level. This contians the OS partition aswell now I need to replace the disks with 300GB SAS disk with 15000rpm speed.
Please let me know the possiblites how do i keep my server alive with out reinstallation.
Please let me know the possiblites how do i keep my server alive with out reinstallation.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
dlethe you can always restart the system by starting it from the HDD you pulled out (let's say ID1). That's the way hardware arrays function and, trust me, never ever had a single issue (even with "rolling back" with IDs) in 12 years of thousands of implementation of HPs servers (Compaq --> HP DL, Blades, MLs...)
Agreed, but there is still risk of picking up an unreadable block. Furthermore, when you yank that 2nd disk, the RAID controller will mark it as bad. It can't tell difference between operator yanking a disk and a failure. So now you have left controller in a state where you have to manually teach it that the replacement drive is no good.
Plus, none of this does anything for resizing partition or testing a bare metal backup, but odds are good you won't run into hiccups, just not 0%
Marking bad....not for HP :-)
What if the disk you did not swap decides to die during the rebuild? What if you have an unreadable block? best and safest way to do this is perform a full bare-metal backup, unplug the original drives, build the new RAIDset, format it, then restore.
Not only is there zero risk of data loss, but this gives you a zero-risk mechanism to test a full bare metal restore. If the restore fails, you still have your original RAID set which you can pop back in, and then try again.