Core to scale migration: importing pools takes ~15 minutes

I tried to migrate from CORE 13.0-U6.1 to SCALE 24.04 but I am having an issue with ix-zfs.service taking about 15 minutes to import my pool during boot up.

I tried a fresh install but importing the pools take the same amount of time and then upon each reboot I face the same issue mentioned above.

I have a mixed capacity pool (4x4TB + 2x3TB) connected via HBA.

I submitted a bug to IX systems but I also understand they have their own priorities.

Does anyone have a solution or perhaps a workaround?

Recommendation: While running CORE, if you have not already done so, create a copy of your configuration file.

Next install SCALE 23.10.2 from the ISO, well you could try the upgrade first but if that fails, use the ISO and install fresh. If it boots and reboots file, then restore your configuration file. See if that works better.

No matter what happens, DO NOT UPGRADE the ZFS Feature Flags! This will lock you into the version you have and you will not be able to roll back to CORE.

That’s quite interesting, as I had the same problem upgrading from SCALE 23.10 to 24.04. I’m afraid I don’t have a workaround for you, though–mine was to revert to 23.10.2 and restore all my apps.

Thank you. Thank you. I will attach my 2nd boot SSD, wait for the resilvering and try upgrading to 23.10.2 and reinstall if that fails.

Yup. I have that as well as a full SSD :slight_smile:

Yes. That would be really bad.

Here is the bug report in case you want to follow or add more info. Hopefully it will be checked soon.

Thank you guys. I really appreciate you trying to help me. I will try your suggestions and post back when I have something :slight_smile:

My ticket is at [NAS-128561] - iXsystems TrueNAS Jira, but yours doesn’t have the additional complication of lots of apps like mine does. I’ve dropped a link to yours in mine.

1 Like

Not sure if this is your issue, but if your pool contains a large amount of datasets and snapshots, it can definitely take a while.
Unless TrueNAS is reporting a fault, let it do its thing.

On the next reboot, you shouldn’t experience this issue anymore. , unless your have encrypted datasets and the unlocking will also take some time.

I have 13560 snapshots which I should clean up a bit
But no, it takes a huge amount of time upon every reboot.

13560 isn’t something to be worried about.
I was expecting more like 10x more, at a minimum.

That wasn’t my experience either–at least four separate reboots and two separate attempts at the upgrade had the same result (though it was closer to 20 minutes in my case).

Upgrading to 23.10.2 works \o/

Thank you all for helping. I am really grateful :slight_smile:

I rebooted a couple of times and booting is smooth apart from the 2nd unconnected network interface which takes a while to timeout.

Now to convert my jails into docker containers.

I wonder what kind of black magic 24.4 has…