Help finding out find why a ZFS pool vanished.

[Greg Marsh]

Thank you, Matthew. 

I'm heading to replace some water damaged drywall and will browse through the zpool history output when I get home. 

Cheers,
Greg


Sent from my iPhone

> On Oct 16, 2016, at 08:45, Matthew Seaman <matthew at FreeBSD.org> wrote:
> 
>> On 16/10/2016 01:56, Greg Marsh wrote:
>> A couple days ago, when trying to copy my desktop files to my backup
>> system, I got NFS mount errors. I was quite sore & tired from work (after
>> 25 years of sysadmining, I'm now doing home renovations), so I went with
>> the Windows admin response of 'reboot everything', on both my desktop and
>> server. Then when trying to remount the NFS shares from my FreeBSD system
>> didn't work, I df'd on my FreeBSD server to find the /storage pool was not
>> there.
>> 
>> I was eventually able to get my storage pool back through 'zpool import
>> -af'
> 
> Have you examined the log from the zpool that mysteriously vanished?
> 
>  # zpool history -il poolname
> 
> This will record every zfs(8) or zpool(8) command affecting the pool.
> You should also examine any command histories you have for actions
> affecting /boot/zfs/zpool.cache -- while the system nowadays will
> generally be able to find your boot pool without this, it may not be
> able to find /other/ pools.  Lots of recipes for manual intervention
> with ZFS systems rely on manipulating this file, but they frequently do
> not warn against /removing/ the working copy before rebooting, something
> that tends to catch out the tidier amongst us.
> 
>    Cheers,
> 
>    Matthew
> 
> 
>