Try to replace Docker with Podman as suggested by the BOINC wiki:
https://github.com/BOINC/boinc/wiki/Installing-Podman-on-Linux

I had to do quite a bit extra to get it working with systemd, it was not so hard to get it working with cgroupv2 but this is not supported I was told, I documented my results on the BOINC Q&A

https://github.com/BOINC/boinc/discussions/6833

In reply to computezrmle's message of 11 Aug 2026:

If your computer has enough free RAM consider to spend a bit more per VM.
This snippet shows how and where "memory_size_mb" needs to be placed:
<app_config>
<app_version>
<app_name>Theory</app_name>
<cmdline>--memory_size_mb 2000</cmdline>
</app_version>
</app_config>
I am trying this app_config:
<app_config> <project_max_concurrent>2</project_max_concurrent> <app_version> <app_name>Theory</app_name> <plan_class>vbox64_theory</plan_class> <cmdline>--memory_size_mb 2048</cmdline> </app_version> </app_config>
I will test if this will be a cure for the nlo-pw-dipole "going crazy". :)

I don't there is anything we can do when you see HTTP gateway timeout. its the backend infrastructure overloaded, it normally comes back eventually

Try Podman instead of Docker.

Further advice can't be given as long as your computers are hidden.
You may make them visible so other volunteers have access to the logs.
https://lhcathome.cern.ch/lhcathome/prefs.php?subset=project

Nothing wrong with your setup, there is an issue with the way the wrapper parses the name of the container when they have long names.

its in this thread. https://lhcathome.cern.ch/lhcathome/forum_thread.php?id=6501

OK I'll see that, I'm asking because projects may have a compatible application, and not generate any task for it (I've seen this in the past), I think that task generation is not automatic or by default for "any OS with a declared app" and requires some configuration.

I'm running the nightly one and its good, they didn't change the code the way I did but same out comes.

https://lhcathome.cern.ch/lhcathome/results.php?hostid=9930009&offset=0&show_names=0&state=4&appid=13

I guess there should be a request for a v23

In reply to Toby Broom's message of 4 Jun 2026:
I have an overide in appconfig for the VM that it has 1.5 GB to aviod this, I think its just a bad WU?
OK. That might be the case.

In reply to Pascal's message of 29 May 2026:
... could you give me a step-by-step description. ...
My previous post describes the steps you need to do.

For example #3:
Use an editor you are familiar with and modify /etc/cvmfs/default.local as follows:
- add 'alice.cern.ch' to the repository list
- remove 'geant4.cern.ch'

For example #4:
Use an editor you are familiar with and modify /etc/cvmfs/default.local as follows:
Add 'CVMFS_USE_CDN=yes'

Similar with the other numbers.
I will not explain how to create a file, open a file in an arbitrary text editor, enter a line, modify some text, remove another line, save the file ...

... if I understood correctly, you have to install cvmfs + podman and that’s it.
Is the installation of CVMFS for podman identical to the one used to do the theory native tasks?
The docker/podman app replaced the older Theory native app.
Podman (or docker) is a MUST HAVE to run this app.

A local CVMFS is recommended as it avoids a lot of network traffic but it is independent from Podman.
If you don't install it, a CVMFS inside the container will be used.
If you decide to use CVMFS on the host it MUST be correctly configured!
As for CVMFS you got the links with step by step instructions and comments in your task logs.

Done its https://github.com/BOINC/boinc/issues/7100

I used to get very long Theory work units, such as some more than 12 hours (on a 2 year old CPU) such as sent dates April 5, 13, 30 and May 1 and 8 (such as https://lhcathome.cern.ch/lhcathome/result.php?resultid=435330784). Some long tasks included those with Input file metadata indicating pythia8, sherpa, herwig, powheg-box, and likely other subtypes.

Since May 8, all work units complete in usually under an hour and a few within about 3-4 hours on that same PC system,

Has the generation of the long work units changed or is it just random over time?

Just wondering whether we still need to request a red message in the Event Log to warn users when a specific task may take a very long time to run.

Sadly, that's not the same problem. It completed on my end successfully and the task has online the status 'Completed, waiting for validation' but it has been waiting for 10 days.
Some tasks had similar results, but they got sent out again to another user, and when that was done, it finished up.
But here it seems that the sending-out part was not generated successfully.

I don't have htop inside the container so I just used top e.g.

podman exec -it ed746bb7ffd4 top -n1

Tasks: 17 total, 2 running, 15 sleeping, 0 stopped, 0 zombie %Cpu(s): 0.4 us, 24.3 sy, 72.5 ni, 2.9 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st MiB Mem : 511227.9 total, 447688.1 free, 53287.8 used, 17278.2 buff/cache MiB Swap: 8192.0 total, 8192.0 free, 0.0 used. 457940.1 avail Mem PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 965 boinc 30 10 635044 269404 14120 R 100.0 0.1 3196:45 Herwig

Since I have more than 1 container I used podman ps to get the ID (ed746bb7ffd4 ) that I was intrested in.

Very cool. I had a bot modify it to run in wsl.
I am attached to 2 very slow huge tasks at the moment.
Both are over 26 hrs old and still chewing.