I have a handful of Docker containers that have stopped updating as of a few weeks ago. When I go to apply the update, there are no obvious points of failure - the old container is stopped and removed successfully, and the new one installed. But the "update ready" prompt remains.

 

This is happening on a few containers. I found one solution was to update the DNS to use Google's DNS servers, so I tried that. I also tried removing the container manually and reinstalling it from previously installed apps. I was able to get a few of the containers to work properly after these two steps, but there are a couple that even when removing and reinstalling, just refuse to update.

 

All three are containers from Linuxserver.io.

• Calibre
• Calibre-web
• Overseerr

 

Any advice would be appreciated. Thanks in advance.

excelsior-diagnostics-20231023-0521.zip

Sorry to bump this, but I am still looking for a solution. Does anyone have any ideas I may not have already tried?

UPDATE: I decided to go a step further with the DNS settings and updated my router's DNS to Google's DNS servers. So, now both Unraid and my router have Google's DNS.

 

This seems to have fixed the issue. There was one docker container I needed to delete and reinstall, but the process is fairly quick and didn't require me to set everything up from scratch. Outside of that, so far so good.

 

I'll update this thread if I run into more problems.

On 11/4/2023 at 1:00 AM, PsionStorm said:

UPDATE: I decided to go a step further with the DNS settings and updated my router's DNS to Google's DNS servers. So, now both Unraid and my router have Google's DNS.

 

This seems to have fixed the issue. There was one docker container I needed to delete and reinstall, but the process is fairly quick and didn't require me to set everything up from scratch. Outside of that, so far so good.

 

I'll update this thread if I run into more problems.

sadly this still isn't working for me and I can't find a solution...

 

I used Pihole for DNS, which used Quad9. I changed DNS back over to my router that uses 8.8.8.8. tried updating the docker images, still nothing. I've tried removing the images and re-adding them from templates. nothing. I have plenty of space on my docker image....idk i'm kinda at a loss here.

7 hours ago, javinp said:

sadly this still isn't working for me and I can't find a solution...

 

I used Pihole for DNS, which used Quad9. I changed DNS back over to my router that uses 8.8.8.8. tried updating the docker images, still nothing. I've tried removing the images and re-adding them from templates. nothing. I have plenty of space on my docker image....idk i'm kinda at a loss here.

My fix seems to have been temporary anyhow. Some containers are no longer updating again.

Is the issue still happening for you?  I find it hard to believe that Unraid would want to force everyone to use 8.8.8.8 as their dns resolver.  I use my gateway, which then points to Cloudflare on 1.1.1.1.  It would be nice to hear someone from the team chime in if this is a known bug or something they may be tracking.

Unfortunately, many routers for some weird unknown issues have problems resolving requests from clients.  This can result in what you're seeing.

 

Unraid does not force anyone to use Google's DNS (8.8.8.8 / 8.8.4.4) as the resolver.  You can set the addresses in Network Settings to be anything you want (I'm a fan of OpenDNS - 208.67.222.222 / 208.67.220.220) including CloudFare (1.1.1.1)

 

Another issue that does come up from time to time is that on occasion (this seems to lately happen to ghcr repositories) is that the checksums reported by ghcr when checking for updates are actually incorrect.  This results in ghcr telliing Unraid / Docker that an update is available, but there may not actually be one (Resulting in a 0 byte pull).  This issue is purely GitHub's (owned by Microsoft) and nothing to do with Unraid at all.

 

 

I think I solved my issue, my system share was set to use the Array instead of Cache.  Once I switched the system share to use Cache, the containers updated and marked themselves as updated with no issues.

I'm having this issue also

 

I noticed a week or so ago that a few containers permanently said "apply update"
I applied the update, everything seemed to go fine, but then they would still say "apply update"

 

 

it certainly seems to be updating images - and after the docker run command all completes succesfully

 

 

UI still shows an update pending

 

if I re-run the update, it pulls the same 5 updated images, reports success - and rinse and repeat

 

this is growing now - it was originally readarr and calibre, now qbittorrentvpn and mediaelch too

 

I've also tried installing the update from the apps page - which does the exact same thing, downloads 5 new images, reports success, still shows update required.

 

My DNS has been set to cloudflare always, but this obviously isn't a DNS issue as it resolves, and isn't a bad checksum issue as there are updated images to pull.. Not sure if there's some other logs I could check somewhere?

Edited January 18, 20242 yr by PhilBarker

I thought as an experiment I'd try removing one of the apps stuck on update, removing the image, and then re-install

 

This seemed to work except now

 

 

Readarr I think updated but now updates are "not-available"

Phil Barker if you look at the screenshot you posted updating qbittorrentvpn you will notice that the last line is extracting a 266mb image and it never gets to PULL COMPLETE (like as if the task has a timeout too short) and continues on to finish the whole task even though the last image never completed extracted.

 

This is where the problem is and I have several containers with the same problem and it always seems to happen if one of the image lines has something over 200+ mb that it needs to extract!!

 

Anyone know how to solve this?

I am seeing the same with home assistant as shown below

I find that sometimes after a few automated attempts the update goes through but specifically with home assistant I still have issues. Trying to look into solutions when I have time.

I have similar problem. About 1-2 hours after updating container it usually comes up with notification that update is ready. When I update it again it pulls 0 bytes and recreates container. Its pretty annoying. I'm using Uniraid 6.12.8 and I remember seeing this problem back in 2023.

Edited March 21, 20242 yr by Antoni Żabiełowicz

The only way I have found to update containers that fail to complete the update directly from UNRAID is to install the portainer container and use that to update the problematic containers.

 

(Inside portainer, container, find the one you want to update and choose recreate, selecting pull latest image.)

 

Once portainer completes its job for the container in question, go back to update in UNRAID (still showing as needing to update) and run it for the container, since it doesn't have to download anything it finalises the update marking it up-to-date.

 

 

I have/had this same issue. For me its only Home Assistant and none of my .arr containers.

-I updated HA 3 times, couldn't notice anything failing.

-Removed it along with deleting the image.

-Reinstalled and pulled with the saved template, it took a longgg time to pull, and it didn't run normally. When going to start the new container, it said local image not found, then below that went to pull some more, but now line by line. It finally finished and then all the text that was out of place disappeared, and it looked like a normal install again.

-Then it gave the update not available under status, with the broken link symbol.

-Ran check all containers for updates, and now it's (knock on wood) back to normal. Says up to date with green check mark.

Just wanted to bump this conversation and see if anyone else is/was having this issue recently and found a solution or root cause?

 

In the last couple of days, while trying to update some containers through the 'action center', all seems to proceed as usual, albeit much slower than before, but once done, the update persists and doesn't actually get updated. The two specifically that I am struggling with right now are; Jellyfin and Jellyseer. I did have a couple of others but repeated attempts got it to go and one (netdata), I was able to update from within docker itself.

 

I am running 6.12.10, the 'action center' did have an update last week and the only other thing I have changed was swapped out a cache pool drive, though, everything is working as I'd expect.

Hi CountryBoy_71,

 

I am still having this issue and though I still haven't found the root cause, I believe its due to a timeout when extracting image files larger than 200MB.

 

The only workaround I have found is as I mentioned above whereby you install portainer, and use that to update the problematic containers, before going back and running the update in the Unraid GUI which is only to change the status to up to date (since it no longer has to download the images and uncompress them).

 

I have been having to use this method for the last 2-3 months to update certain containers.

 

How do we bring this to the attention of the devs?

 

 

Seems like this has been a problem for a while with containers that have images over 200mb, see this comment;

 

 

Hi F1Spain, 

 

Thanks for the information...I am wondering, if that is the case, is there any way to manually download and extract...in effect, manually updating these?  Right now, it's not a huge deal for me as the two that aren't updating are rarely used but I know, soon, ones I do will have an update and I'm sure they'll be over to 200 mark!

Any headway on this matter?  I'm started having the issue this morning with a couple dockers.  I decided to try to use "Check For Updates" at the bottom of the docker screen.  Now ALL of my dockers are saying "not available".  😞

 

Following...

I have had similar problems, and as stated usually involving updating containers with larger files. Seems it would partially thru downloading file and then stall.  My docker.img file was located on a ZFS formatted drive. Removing and recreating did not solve this for me.  What did worked was changing from .img to directory under docker settings

setting -> docker -> enable docker = no -> change from vDisk to Directory, then reenable.

Mine just started working out of the blue, no changes on my end.  

I solved my issuse with  Nextcloud and Immich not updating this way:

 

Go to settings

Select docker

Change "Docker Stop Timeout:" to 120 seconds.

 

Issue gone  🙂 (Please see update below screenshot)

Unraid version 6.12.10

 

Update. It did only work once, sorry  😞

Today the plugin Fix Common Problems had a warning that my TRIM schedule was faulty, and that i needed to disable and then enable it to get it working correctly. I did that and docker update worked again.

 

I'm not going to write solved this time  🙂 but i worked and I updated Nextcloud and Immich without any issues.

Edited July 14, 20241 yr by Bonzoom

On 3/21/2024 at 6:18 PM, f1spain said:

The only way I have found to update containers that fail to complete the update directly from UNRAID is to install the portainer container and use that to update the problematic containers.

i went and looked in the apps tab and i only found Portainer. (Portainer-Agent, Portainer-BE, PortainerCE, PortainerCi) which one are you using?

 

It seems the portainer I have installed is no longer updated, the one to use would be PortainerCE.