Since upgrading to 7.2.0 I've been having an intermittent write issue with one of my shares. I'm not 100% confident that it was caused by upgrading to 7.2.0, but it started happening the day after. I have a fix/workaround so wanted to share it here in case other folks run into it.

I have a share which is set up with a cache pool and moves to my array overnight. After the mover completes I can no longer write to the share at /mnt/user/Share/... and I get an Invalid argument error when I try to do so.

touch /mnt/user/Share/test
# touch: cannot touch '/mnt/user/Share/test': Invalid argument

However I can still write to the array directly at /mnt/user0/Share and I can also write to the cache pool at /mnt/cache. I have no SMART or filesystem errors. The cache is a ZFS mirror.

If I stop/start the array everything works as expected until the next night when the mover runs.

When investigating the issue I noticed that the directory /mnt/cache/Share didn't exist. I'm not sure if it did exist before upgrading to 7.2.0, or if it did exist before the mover ran.

My fix was to manually create that directory

mkdir /mnt/cache/Share

My suspicion is that the mover cleaned up this directory, and it needs to exist in order to write to the share through the cache. I also am guessing that stopping/starting the array caused this directory to be recreated. However, so far things have been working since I manually created the directory and remain working 24 hours later after the mover has been run.

It's a known issue for some users with zfs as primary storage; it can happen after the mover runs and it should be fixed for 7.2.1, as a workaround, typing zfs mount -a should resolve the it.

Thanks for the fast response! When you say primary storage do you mean the array? My main array is using unraid, and my cache pool is using ZFS.

With a share set to use a zfs cache pool as the primary storage, array (or another pool) as secondary.

Got it, much appreciated!

On 11/5/2025 at 10:36 AM, JorgeB said:

It's a known issue for some users with zfs as primary storage; it can happen after the mover runs and it should be fixed for 7.2.1, as a workaround, typing zfs mount -a should resolve the it.

I was affected by this issue as well. This workaround worked for me.

Any ETA on 7.2.1?

The last info I have is that LT was planning to release a 7.2.1-rc.1 with this and some other fixes this week. This would allow the affected users to confirm it resolves the issue, since I and LT cannot reproduce it, and not for a lack of trying on my part, and if all looks good, then release 7.2.1 stable a week or so later.

Can confirm the problem came back overnight and zfs mount -a resolved it. I don't usually test pre-releases but if you want to ping this thread when 7.2.1-rc.1 is available I'm happy to give it a go and feed back if it resolves things.

Edited November 7, 2025Nov 7 by jacobtomlinson

5 minutes ago, jacobtomlinson said:

if you want to ping this thread when 7.2.1-rc.1 is available

I will.

7.2.1-rc.1 is out, and it should resolve this issue, please retest.

18 hours ago, JorgeB said:

7.2.1-rc.1 is out, and it should resolve this issue, please retest.

I am unable to reproduce the issue on 7.2.0 by generating data and manually running the mover.

Regardless I upgraded to 7.2.1-rc.1. I will monitor to see if the issue occurs again.

Hopefully @jacobtomlinson has better luck in reproducing the issue on 7.2.0 and can confirm if it is resolved in 7.2.1-rc.1.