Worrying looking errors!!!!

[b0redom]

Sorry for the slow response on this one. I've went away on holiday and turned it off to ensure no further data loss.

I was under the assumption that like Linux etc disk enumeration doesn't necessarily persist across reboots. So what could happen is:

sata0 - ada0
sata1 - ada1
sata2 - ada2
sata3 - ada3

Disk sata1 dies for whatever reason, lets say the sata cable is faulty, so this happens:

sata0 - ada0
sata1 - dead
sata2 - ada1
sata3 - ada2

I know this is how it work as I actually had a duff SATA cable when using Openfiler. It caused a LOT of headaches as when sata1 reappeared, it would become ada3 etc.

For this reason I used smartctl to dump out the serial number of the failing disk. I could of course have made a mistake, but I'm pretty sure I swapped out the right disk.

Can someone confirms that disk enumeration works like this, and not like Solaris where disk enumeration is always c0t0d1, c0t0d2 etc depending on what channel the disk is connected to?

Also is there a guide to replacing a busted disk for the future?

Cheers....

b0redom

 

[ProtoSD]

Also is there a guide to replacing a busted disk for the future?

The Unofficial FAQ in my signature below.

 

[b0redom]

Thanks for that! Can anyone answer the disk enumeration question? I guess that's BSD specific rather than Freenas though.

Regards...

b0redom

 

[b0redom]

So, I've reinstalled my FreeNAS. Turns out I did pull the wrong disk!

I'm still, however, seeing smartd returning:

/dev/ada0, 1 currently unreadable (pending) sectors
/dev/ada3, 1 currently unreadable (pending) sectors

Using a Windows 7 box I have, I ran full 'extended' checks using the Western Digital drive checking tool, and all (except the bad disk) passed with 0 SMART errors. As soon as I put them back into FreeNAS, they're showing up again.

Is there something I can do to force the drives to fail so I can RMA them? I really don't trust them having had one go pop.