Hello I'm using Unraid 7.2.0. Whenever the docker service is started while the array is running or before being started and then started. The webpage timesout and the server is unable to be reached. The VM manager is able to be enabled just fine. This began when I changed the PCIe ACS override from disabled to downstream. My original goal was to break up some IOMMU groups.

So far I've tried some troubleshooting in order below:
- Booted into Safe GUI mode

- Cleared out the "/boot/config/plugins/dockerMan/templates-user/" directory.

- Renamed savedpcidata.json to savedpcidata.old to regenerate file.
- Deleted docker directory(I use dir instead of an image).

I've attached my diagnostic zip and my previous syslog zip.
Appreciate the help.

powerdedication-diagnostics-20251122-0403.zip powerdedication-syslog-previous-20251122-1102.zip

Edited November 22, 2025Nov 22 by ShredderCody

Does it still happen if you disable the VM service?

Just now, JorgeB said:

Does it still happen if you disable the VM service?

Yes, disabling VM service and starting the array with docker service enabled causes the hang/timeout.

I don't see any obvious issues logged, since it's likely a config issue, you can try redoing the flash drive, backup the current one first and then recreate it using the USB tool and just restore the bare minimum, like the key, super.dat and the pools folder for the assignments, also copy the docker user templates folder (\config\plugins\dockerMan\templates-user), if all works you can then reconfigure the server or try restoring a few config files at a time from the backup to see if you can find the culprit.

3 hours ago, JorgeB said:

I don't see any obvious issues logged, since it's likely a config issue, you can try redoing the flash drive, backup the current one first and then recreate it using the USB tool and just restore the bare minimum, like the key, super.dat and the pools folder for the assignments, also copy the docker user templates folder (\config\plugins\dockerMan\templates-user), if all works you can then reconfigure the server or try restoring a few config files at a time from the backup to see if you can find the culprit.

I redid the flash and that did restore my server to functional, this time I did use efi, as opposed to legacy. I was even able to accomplish my original goal of the enabling the acs override and the issue hasn't returned so far. So your speculation of a config issue is probably right. No idea what caused it and I doubt the efi switch prevented the issue from returning. Thanks.

Hello All,

I would like to report that I have this issue also. Fortunately I have a 3rd party docker manager managing my containers (Komodo) so that they don't startup when the array starts up.

After spending the day root causing the issue I find that the webGUI for Unraid becomes unresponsive if two docker containers are sharing an NVIDIA GPU via the docker nvidia toolkit.

I anticipate this issue is manifesting across the board with people running HW acceleration for their plex/jellyfin containers and containers dedicated to transcoding or GPU compute tasks.

I will continue to investigate.

Only other things to note so far is that the unraid server is unreachable via webGUI but is REACHABLE with ssh and through my Komodo control console. stopping containers using the GPU does not bring back the webGUI.

11 hours ago, trimatrix said:

Hello All,

I would like to report that I have this issue also. Fortunately I have a 3rd party docker manager managing my containers (Komodo) so that they don't startup when the array starts up.

After spending the day root causing the issue I find that the webGUI for Unraid becomes unresponsive if two docker containers are sharing an NVIDIA GPU via the docker nvidia toolkit.

I anticipate this issue is manifesting across the board with people running HW acceleration for their plex/jellyfin containers and containers dedicated to transcoding or GPU compute tasks.

I will continue to investigate.

Only other things to note so far is that the unraid server is unreachable via webGUI but is REACHABLE with ssh and through my Komodo control console. stopping containers using the GPU does not bring back the webGUI.

11 hours ago, trimatrix said:

Hello All,

I would like to report that I have this issue also. Fortunately I have a 3rd party docker manager managing my containers (Komodo) so that they don't startup when the array starts up.

After spending the day root causing the issue I find that the webGUI for Unraid becomes unresponsive if two docker containers are sharing an NVIDIA GPU via the docker nvidia toolkit.

I anticipate this issue is manifesting across the board with people running HW acceleration for their plex/jellyfin containers and containers dedicated to transcoding or GPU compute tasks.

I will continue to investigate.

Only other things to note so far is that the unraid server is unreachable via webGUI but is REACHABLE with ssh and through my Komodo control console. stopping containers using the GPU does not bring back the webGUI.

In my case networking was completely dead. SSH would lose connection and the server would not respond to pings. I did have some containers using gpus set using "all".