SOLVED PANIC, CRITICAL: db> after trying to remove snapshots

[Dubravko Vukmirovic]

Hi
First thank you reading for this thread and thank you so much for any advice. My heart is in pain because this is personal storage with family movies, pictures and study material.

It was running for years smooth so I need to pull from my memory info:
4 hard drives each 2 TB
RAID 10
8 GB

It started with having free 150 GB but could not copy one file
Found 489x389MB auto snapshots
stopped replication and snapshot, rebooted
run
zfs list -H -o name -t snapshot | xargs -n1 zfs destroy

[root@x] ~# zfs list -H -o name -t snapshot | xargs -n1 zfs destroy
cannot destroy 'Jails/Plugins/.warden-template-pluginjail@clean': snapshot has dependent clones
use '-R' to destroy the following datasets:
Jails/Plugins/crashplan_1
cannot destroy snapshot RAID10@auto-20150930.1700-2w: dataset is busy

I run
zfs list -H -o name -t snapshot | xargs -n1 zfs destroy -R

and I got
db>

I have only console access now

Please see screenshot

 

[Robert Trevellyan]

Looks like you're using hardware RAID. That reduces your chances of a successful outcome. Perhaps you destroyed your boot volume.

1. Take a new USB stick
2. Do a clean install of the same version of FreeNAS that you were running before.
3. Restore your saved configuration, if you have one.
4. See if you can import the volume.
5. Report back.

 

[depasseg]

The safest thing you can do is to just restore from your replicated data. By deleting the snapshots using the CLI like you used, you removed system snapshots as well as you data snapshots.

 

[Dubravko Vukmirovic]

3 hours of stress and I am up.
I will post my experience and improvements soon.

In short:

1. Take a new USB stick
2. Did a clean install of the same version of FreeNAS that you were running before.
3. Restore your saved configuration
4. All volumes recognized automatically
5. Reporting back :)

There is more but later...