Talk about bad luck....I had a drive red-ball on me shortly after upgrading from 4.7 to 5.0-rc3. Ran parity check twice before upgrading (no errors) but the first parity check after upgrading I get a red-ball.

 

The upgrade process completed successfully from the looks of it. Array started without error, all drives green, ran the permissions script, etc. Problem began when I ran the check parity with NOCORRECT option to exercise the array a bit. I'm sure it is a bad drive. The syslog has "media error" entries for failed drive and smart report shows ~900 sectors pending relocation.

 

Anyways, the failed drive & parity are both 2TB. I could buy a new 2TB and replace the failed drive but I'd like to move to 3TB drives as they're more economical. The swap-disable procedure sounds ideal but the lack of posts on that subject leads me to believe it's a rather uncommon/untested procedure.

 

What do you guys think? Should I try swap-disable or just buy a new 2TB?

 

 

If you have enough room left on the other drives in the array, or on your other computers, I'd copy the data off the virtual array failed drive to another physical drive somewhere in case the procedure doesn't work. If you have no backups of important data, I'd concentrate on getting the array protected and or backed up first before trying a parity swap upgrade. If the data is replaceable, go for it. Keep in mind you risk total data loss on two drives (or more) if something goes badly wrong.

 

It'll most likely work, but I wouldn't try it on my data.

 

As a compromise, you could HPA short stroke the 3TB to match your existing 2TB, then when you can afford another 3TB, install it as parity, then remove the HPA and presto after rebuilding, another 1TB of space. You wouldn't be gaining the extra TB right now anyways, so no difference there. This is what I would do facing your dilemma.

Thanks for the advice. I like your approach better than the swap-disable option so I'll go with that.