Dell 1950 RAID 1 Drives.

Tim Judd tajudd at
Thu Jan 1 00:32:02 UTC 2009

Keith wrote:
> Have the above server, drive 0 failed. No onsite spares left. Ordered
> replacement drives a week ago but they are still on route and wont be here
> till Jan 5 or 6.
> All I have is a test server of the same make/model in my test rack
> configured and running a slightly different version of FBSD.
> If I pull a drive from this test server that is already configured to a
> RAID contoller and put it into the server with the bad drive, how will
> the machine deal with it as it already has a valid config on it? Will I
> just be able to go to the RAID Bios and rebuild?
> Or wipe the config on the test server so the drive has no config on it
> before installing it into the server with the failed drive?
> Are the SAS drives in a Dell 1950 hot swap by chance?
> Have not had a drive go like this with no spares around before.
> Thanks.

Better to ask Dell weather they're hot swap or not as an absolute 
answer.  When I asked, the tech at the time said "the newer PERC 
controllers are all hotswap, as long as the RAID controller isn't 
talking to the drive anymore".  Given the 2nd system is actively using 
it, I don't think it's hotswap.

As for the config, you'd have to force an integrity/rebuild on the 
first.  Clearing the config on the 2nd shouldn't be necessary.  In fact, 
the second may see a missing drive and panic as is.  The PERC6 
controllers I deal with at work have a "feature" that I would call a 
bug.  A missing physical disk causes the RAID controller to 
waste/dump/ignore/forget it's onboard config.  You may be dead in the 
water on the 2nd system until you get a drive to put in.

Backup server 1 and 2.  And wait for the drive to come.  No other good 
way to solve this problem.

More information about the freebsd-questions mailing list