Apps Fail to Start Running After Update

r0b0ty · July 13, 2025, 12:01am

Hi, guys. I’m running TrueNAS Scale 25.04.1. Yesterday, two of my apps showed Updates Available. I proceeded to update the first one and got an error I had never seen. I took a chance and tried the second app update, and got a similar error. Obviously not a coincidence. Does anyone know what the error means and how I can fix this? I tried to Roll Back to my previous version, but it requires that the app be running, which it obviously can’t… so I’m stuck.

This is the error for one of the apps (identical to the second app):

And this is the content of the app_lifecycle.log file:

[2025/07/11 18:42:20] (ERROR) app_lifecycle.compose_action():56 - Failed 'up' action for 'paperless-ngx' app:  tika Pulling \n permissions Pulling \n postgres_upgrade Pulling \n gotenberg Pulling \n redis Pulling \n paperless Pulling \n postgres Pulling \n paperless Error Get "https://registry-1.docker.io/v2/": dial tcp: lookup registry-1.docker.io on 10.0.25.1:53
: read udp 10.0.25.4:58863->10.0.25.1:53: i/o timeout\n postgres Error context canceled\n tika Error context canceled\n postgres_upgrade Error context canceled\n permissions Error context canceled\n redis Error context canceled\n gotenberg Error context canceled\nError response from daemon: Get "https://registry-1.docker.io/v2/": dial tcp: lookup registry-1.docker
.io on 10.0.25.1:53: read udp 10.0.25.4:58863->10.0.25.1:53: i/o timeout\n
[2025/07/11 18:43:07] (ERROR) app_lifecycle.compose_action():56 - Failed 'up' action for 'paperless-ngx' app:  tika Pulling \n redis Pulling \n tika Error Get "https://registry-1.docker.io/v2/": dial tcp: lookup registry-1.docker.io on 10.0.25.1:53: read udp 10.0.25.4:40191->10.0.25.1:53: i/o timeout\n redis Error context canceled\nError response from daemon: Get 
"https://registry-1.docker.io/v2/": dial tcp: lookup registry-1.docker.io on 10.0.25.1:53: read udp 10.0.25.4:40191->10.0.25.1:53: i/o timeout\n
[2025/07/11 18:44:19] (ERROR) app_lifecycle.compose_action():56 - Failed 'up' action for 'syncthing' app:  syncthing Pulling \n syncthing Error Get "https://registry-1.docker.io/v2/": dial tcp: lookup registry-1.docker.io on 10.0.25.1:53: read udp 10.0.25.4:54316->10.0.25.1:53: i/o timeout\nError response from daemon: Get "https://registry-1.docker.io/v2/": dial t
cp: lookup registry-1.docker.io on 10.0.25.1:53: read udp 10.0.25.4:54316->10.0.25.1:53: i/o timeout\n
[2025/07/11 18:58:53] (ERROR) app_lifecycle.compose_action():56 - Failed 'up' action for 'syncthing' app:  syncthing Pulling \n syncthing Error Get "https://registry-1.docker.io/v2/": dial tcp: lookup registry-1.docker.io on 10.0.25.1:53: read udp 10.0.25.4:42226->10.0.25.1:53: i/o timeout\nError response from daemon: Get "https://registry-1.docker.io/v2/": dial t
cp: lookup registry-1.docker.io on 10.0.25.1:53: read udp 10.0.25.4:42226->10.0.25.1:53: i/o timeout\n

Thanks for your time!

Astonmid · July 13, 2025, 3:56am

Have just encountered the same issue, so clearly not something unique to your setup. I’m also looking for a solution.

architeck · July 19, 2025, 8:06pm

Same for me. Was able to rollback three of my apps and they work again.

r0b0ty · July 19, 2025, 9:36pm

How were you able to roll them back? Were they still running after your update? When I tried, I got an error stating that the app had to be running to do a roll back. Interesting…

Astonmid · July 20, 2025, 5:37am

I have always encountered patchy response to “updates” since scale went from Kubernetes to docker. Several have failed in the past, but I have always been able to start them so can either try again later (if the fail was to reject the new version entirely) or roll back if the update appeared to have worked, but the thing doesn’t seem to respond (as in trying to get into the GUI.)
This time the updates seem to stall mid-course, can’t roll back, can’t go forward. I have all my config files stored separately (don’t use “ix-volumes”)
I tried deletely the app and reinstalling, but TrueNas said I was already using that name. Tried a new name and still didn’t work. In the end I deleted all my config files, generated new ones and rebuilt the apps from Scratch. All my apps are now working. Lastly (this is the one I feared updating the most) a few days later I updated NGINX. Took a very long time, particularly deploying, but it updated on its own.
Don’t know if any of this helps anybody!!

r0b0ty · July 20, 2025, 9:59am

Thanks @Astonmid. I was planning on the uninstall-reinstall approach as a last resort, and hoping to use the old config files too. Seems like I may not be able to reuse them, but I’ll try. Worst case, I’ll start from scratch as you did. Was really trying to avoid that. Again, thanks for the input.

architeck · July 21, 2025, 1:28am

They were stopped and it still let me rollback. I didn’t do anything special? I’m on Electric Eel 24.10.2.2

architeck · July 24, 2025, 2:35am

I just updated to 24.10.2.3 and I think coincidently or not, the apps updated and started up fine. But I’m getting other buggy things like connection issues.

I did a fresh reinstallation of Electric Eel, and then decided to just upgrade to Fangtooth and everything seems okay …so far…

r0b0ty · July 26, 2025, 3:52pm

The problem resolved itself!
I just wanted to update everyone that the apps with the original problem now run without errors. Although I’m pleased, it’s a little annoying because I did nothing except wait (over a week) and retry. Today, all I did is click on the play button to run the apps… and they did without errors. I did not roll back to a previous version and I did not update TrueNAS (there was none). We did have a power outage at some point, during a thunderstorm, but I also had rebooted my TrueNAS machine a couple of times when I was experiencing the issue… with the problem still present after the reboot. So I don’t think the power outage had any impact.

Thank you @Astonmid and @architeck for your feedback!

Astonmid · July 26, 2025, 11:22pm

Yes - I’ve used a similar technique successfully in the past. After having a “fail”, I’ve left it overnight/several days and when I tried it again it sailed through effortlessly. Just thought it was community nature of the system and possibly the developers had done a twerk or two between my two tries, but no evidence or knowledge to base this assumption.

manna · July 27, 2025, 3:08am

I know that this problem already fixed itself, but just if you are interested, it looks like this was an issue with your Truenas box not able to connect to the website to download the updated app.