Drive is Degraded, what do I do now?

[awildthingy]

Hello all,

So Ive been lurking around here for a while but this is my first cry for help. I have wanted to setup a Freenas system for forever and 2 months ago I scored a used Supermicro Server for a great price, The purchase came with 8 old but working drives; despite this I sought to get a few new drives just in case. On 4/1/17, the first drive failed - I was able to replace it and resilver without a problem; I'm not sure what version of Freenas I was using at that time but I have been pretty on top of installing new stable releases as they have become available. I'm currently running Freenas 10.0.4 (stable), and my second drive failed yesterday. I was able to offline and replace that disk, but immediately after my topology looks like first screenshot. To remedy I have restarted the machine and I have looked at the CLI for more info on what is going on, then tried to offline the disk (by process of elimination, I fugure we are dealing with "da2") (see second screenshot). I assume this is another drive failure but I am not sure what the next step is. I cannot seem to offline this thing; would it be ok in this case to pull the drive out of the HBA? A separate but related question is there a a log file that is compiled by the system in a default setup that I can access and could shed more light on this? If so, how do I get to it? Overall, any advise would be appreciated.

P.S. - No there is not crazy important data on this thing yet, Everything is backed up.i So it would only be a standard sized hassle to start over but I would rather avoid that if I can. Thanks

 

[nojohnny101]

Hello,
Have you seen this?
https://forums.freenas.org/index.php?threads/important-announcement-regarding-freenas-corral.53502/

Corral had poor support for replacing drives (non-existent in the GUI to my knowledge) and it had to be done via zfs commands which is never recommended (but in corral's case, was the only way).

If I were you, I would start over and install the latest release of 9.10.2 They are releasing an update to 9.10 on Monday (05/01/17) which will essentially be 9.10.3 but they are calling it FreeNAS 11.

 

[awildthingy]

Oh man, no, I somehow completely missed this. I now see that the freenas.org site offers only 9.10 for download again as well. Well, Ironically I started with 9.10 when I first installed Freenas at the end of February. I agree with you @nojohnny101, given this new (to me) information, I'll need to start over again. As a followup question, the procedure to start over would be to destructively detach the volume, then reinstall 9.10 to the boot drive, right? Or do I actually need to manually wipe each drive to make sure new installation wont pick up old pool?

 

[SweetAndLow]

Corral had poor support for replacing drives (non-existent in the GUI to my knowledge) and it had to be done via zfs commands which is never recommended (but in corral's case, was the only way).

This is incorrect, replacing drives is supported. You use the freenas CLI to do it. This is documented in the wiki. You should never use the zfs commands even in corral.

Sent from my Nexus 5X using Tapatalk

 

[gpsguy]

You shouldn't need to blow away your pool and start over. @Ericloewe has a FAQ that lists the various scenarios and what you need to do.

If you are using Dockers - they aren't supported in 9.10.x yet. And, what was going to be called 9.10.3 will now be named 11.0 (scheduled for release on Monday). It now uses FreeBSD 11 under the hood.

..., I'll need to start over again. As a followup question, the procedure to start over would be to destructively detach the volume, then reinstall 9.10 to the boot drive, right? Or do I actually need to manually wipe each drive to make sure new installation wont pick up old pool?

 

[nojohnny101]

This is incorrect, replacing drives is supported.

Thanks for the correction. What this functionality added to Corral in the latest update before it was killed?

I made that comment based on threads like these that I saw:
https://forums.freenas.org/index.php?threads/unable-to-replace-a-failed-disk-freenas-corral.51644/
https://forums.freenas.org/index.php?threads/freenas-corral-replacing-a-disk-in-a-zpool.52404/
https://forums.freenas.org/index.php?threads/replacing-failed-drives-with-freenas-corral.52242/

Mute point I suppose.

 

[SweetAndLow]

No it was there, no one bothered to look out up though.

Sent from my Nexus 5X using Tapatalk

 

[nojohnny101]

Oh ok.

 

[awildthingy]

Followup: Thanks all, I was lucky, and never had an issue with the drive replacement process in Corral (CLI was broken on this for a version or two). And actually, the weirdly absent drive noted in my original screenshots seemed to recover on its own a few hours after my op. However, I did replace the Corral install with a clean 9.10 install. I didn't touch my pool, I simply imported it into the new boot environment. Pool is now says healthy and I am working on some long smart tests to make sure Im good.

 

[gpsguy]

I am glad to hear that your your system is up and running again.