We had a power outage with our Ixsystems M40 running Truenas 11.3-U3.2. It is on a UPS, but not a generator. We have had power outage before but things always came back online. It is where all of our VMware VMs lives. Upon logging into the unit, the pool “tank” was listed as Unknown. After doing a bit of research, I used the command zpool import -F tank. It gave me an error that the logs device was unavailable and that to us the -m option to bypass that device which I did.
That brought the pool and zvols back, and could see them back in vmware via the ISCSSI (but not as datastores, those were still missing). I still could not access any of the vms, so I restarted the unit. After the restart the pool is now gone. I ran a zpool status and i get the following error “Assertion failed: (reason == ZPOOL_STATUS_OK) file /truenas-releng/freenas/_BE/os/cddl/contrib/opensolaris/cmd/zpool/zpool_main.c, line 5128. zsh abort (core dump)”
If I try to add the pool again it give me a message that the pool already exist (which makes sense)
Try export from the shell, then import from the UI.
A plain zpool import <pool> from the shell will never work, because you need at least the extra parameter -o altroot=/mnt plus whatever else the UI and the middleware might be doing that is not covered by the simple import command.
Please be very careful using the command line, if you are not 100% certain, what it will do. Especially the different import options “-f” vs “-F” for example…
Since you are using iX hardware, you have official support. Wait until someone from iX reaches out to you or replies here in this thread.
I will try an export from the shell and an import from the UI.
Unfortunately, our support expired (an oversight on our part), and they are not willing to help until the support is renewed (the account manager seems to be on vacation) even though we have two other units from them that are in support. The problem is that we are completed dead in the water because that is where AD and all of our management server lives. I do not want to lose this data though.
You can get free support here. Maybe @HoneyBadger can help.
Again, be careful using the CLI.
Thanks for the info. I will reach out to @HoneyBadger
“Assertion Failed” doesn’t seem to line up with “ZPOOL_STATUS_OK” in my mind 
so there might be something more at play here. Can you DM me the serial number of your unit so I can look up what the situation is on the back end?
This might be an issue. Using the -m flag means that the pending writes were discarded, so data that your VMware hosts thought was written was likely there, and the discard operation means it’s been removed. If you’re showing an unavailable log device then we need to see what happened there.
Is this a dual-controller/HA unit?
It turns out that when I exported it before it worked. I rebooted and it errored out because it could not find the log device. The -m option removed that device that was unavailable. iXSystems said it changed and they don’t know why.
We had to remove the log device and then re-added it back. Once that was done and a reboot was performed it worked.