Development for LHC@home

Updated the docker wrapper to v22.

https://github.com/BOINC/boinc/wiki/Docker-wrapper-release-notes

First Task finished with Vers.7.67.

I ran two tasks with this newest docker wrapper.

The first one with the default conditions: verbose level 1 and no network interruption.

https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=3778899

The second one with extended podman logging (verbose level 2) and with no network available at the start.
This causes a waiting for heartbeat condition (without task erroring out) until network is available again.

05 Jun 21:46:04 Suspending network activity - user request
lhcathome-dev 05 Jun 21:46:16 task Theory_2922-4852259-1083_1780390017_0 resumed by user
lhcathome-dev 05 Jun 21:46:17 Starting task Theory_2922-4852259-1083_1780390017_0
05 Jun 21:51:00 Resuming network activity

https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=3778903

In the latter result I did not found the extended logs for podman ps and podman stats

In reply to Crystal Pellet's message of 5 Jun 2026:

In the latter result I did not found the extended logs for podman ps and podman stats

Just to be sure, I retested the logging with verbose values 1 and 2.
There is no difference in the logging. Verbose level 2 should give the extended logging for the podman commands:

ps --all -f "name=etc" and

stats --no-stream --format "{{.CPUPerc}} {{.MemUsage}}" boinc__etc

These loggings were there in the previous docker_wrapper version 18

I don't miss them, because the logging was so extended, that the most important part (building the image) at the beginning of stderr.txt was truncated.

Hope this solve the "neverendig-wus" problem

In reply to Crystal Pellet's message of 6 Jun 2026:

... I don't miss them, because the logging was so extended, that the most important part (building the image) at the beginning of stderr.txt was truncated.

This was the intention behind issue #7001:
https://github.com/BOINC/boinc/issues/7001

If somebody needs runtime stats I suggest to use 'podman stats ...' in a terminal.
It is highly configurable as can be seen here:
https://docs.podman.io/en/stable/markdown/podman-stats.1.html

In reply to boboviz's message of 6 Jun 2026:

Hope this solve the "neverendig-wus" problem

What are "neverendig-wus"?
Especially those where docker_wrapper is the root cause.

In reply to computezrmle's message of 6 Jun 2026:

This was the intention behind issue #7001:
https://github.com/BOINC/boinc/issues/7001

If somebody needs runtime stats I suggest to use 'podman stats ...' in a terminal.
It is highly configurable as can be seen here:
https://docs.podman.io/en/stable/markdown/podman-stats.1.html

Yeah, I know it was planned since the beginning of the use of docker for Theory to reduce these podman command loggings (what is fixed now),
but the verbosity level 2 is not working like intended.

In reply to computezrmle's message of 6 Jun 2026:

What are "neverendig-wus"?
Especially those where docker_wrapper is the root cause.

Docker wus, usually, on my pcs run 2/3 hrs.
Some wus arrived at 100% and still run over 24 hrs.
I called these "neverending"

In reply to boboviz's message of 6 Jun 2026:

In reply to computezrmle's message of 6 Jun 2026:

What are "neverendig-wus"?
Especially those where docker_wrapper is the root cause.

Docker wus, usually, on my pcs run 2/3 hrs.
Some wus arrived at 100% and still run over 24 hrs.
I called these "neverending"

Over @prod it has been explained uncountable times that (and why) you must not rely on BOINC's progress meter.
Instead, check the runRivet.log in slots/n/shared to get an impression what the scientific app is doing.
Depending on the input params a task can easily take several days.

The longer one use herwig7 event generator and I watch them all run and have saved records of close to 75 Valids and many run between 2 days and 3 days

I even got a Sherpa Valid in 15 hours
powheg-box average around 7 - 8 hours
pythia8 over 2 days close to 4 days
Even one that was Run time 22 days 3 hours 2 min 27 sec herwig7

When I see one that stops running and giving them an extra hour to make sure I will abort it but powheg-box are not as easy to watch via runRivet so I just let them run

Just one of them -
Name Theory_2922-4884870-1083_1780382978_1
Workunit 2791113
Created 3 Jun 2026, 14:37:54 UTC
Sent 3 Jun 2026, 16:56:06 UTC
Report deadline 13 Jun 2026, 16:56:06 UTC
Received 5 Jun 2026, 20:22:37 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x00000000)
Computer ID 5326
Run time 2 days 3 hours 25 min 14 sec
CPU time 2 days 0 hours 28 min 19 sec
Priority 0
Validate state Valid - herwig7
Credit 2,110.30
Device peak FLOPS 4.92 GFLOPS
Application version Theory Simulation v7.66 (docker)
windows_x86_64
Peak working set size 13.67 MB
Peak swap size 2.69 MB
Peak disk usage 34.94 MB

+1
Name Theory_2922-4762179-1044_1780155908_1
Arbeitspaket 2789867
Erstellt 1 Jun 2026, 19:07:37 UTC
Gesendet 1 Jun 2026, 19:13:09 UTC
Ablaufdatum 11 Jun 2026, 19:13:09 UTC
Empfangen 6 Jun 2026, 19:01:32 UTC
Serverstatus Abgeschlossen
Resultat Erfolgreich
Clientstatus Fertig
Endstatus 0 (0x00000000)
Computer ID 5116
Laufzeit 4 Tage 23 Stunden 48 min. 11 sek.
CPU Zeit 4 Tage 17 Stunden 51 min. 49 sek.
Priority 0
Prüfungsstatus Gültig
Punkte 8,219.68

Yes Magic,
good thing need some time longer.
powheg (herwig) is not easy to control.

Here is one of my longer herwig7 version units

Name Theory_2922-4893431-696_0
Workunit 2721235
Created 9 Mar 2026, 11:43:01 UTC
Sent 9 Mar 2026, 22:26:15 UTC
Report deadline 19 Mar 2026, 22:26:15 UTC
Received 14 Mar 2026, 10:03:19 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x00000000)
Computer ID 5326
Run time 4 days 11 hours 36 min 47 sec
CPU time 4 days 5 hours 58 min 1 sec
Priority 0
Validate state Valid herwig7
Credit 4,368.90
Device peak FLOPS 4.87 GFLOPS
Application version Theory Simulation v7.66 (docker)
windows_x86_64

On linux, all the docker-theory tasks failed within a few seconds.

https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=3779417

In reply to computezrmle's message of 6 Jun 2026:

Over @prod it has been explained uncountable times that (and why) you must not rely on BOINC's progress meter.
Instead, check the runRivet.log in slots/n/shared to get an impression what the scientific app is doing.
Depending on the input params a task can easily take several days.

I've not seen that thread, sorry.
So, a good news, it's not a problem but "a feature".

So, next is a GPU app!!! :-P

In reply to Laurence CERN's message of 5 Jun 2026:

Updated the docker wrapper to v22.

https://github.com/BOINC/boinc/wiki/Docker-wrapper-release-notes

For me Theory docker (Linux) requires a docker_wrapper that includes pr7113:
https://github.com/BOINC/boinc/pull/7113
To ensure this is included please use an artifact/version from this afternoon or newer.

Older versions can't deal with the longer task ids as reported here:
https://github.com/BOINC/boinc/issues/7111

In reply to Laurence CERN's message of 5 Jun 2026:

Updated the docker wrapper to v22.

https://github.com/BOINC/boinc/wiki/Docker-wrapper-release-notes

What about forwarding the BOINC-Client-Proxy-Setting to the Podman/Docker-App ?

My BOINC-Clients are configured to use my separate running Squid, but the Podman/Docker-Tasks still say:

No local HTTP proxy found.

+1

In reply to Yeti's message of 7 Jun 2026:

In reply to Laurence CERN's message of 5 Jun 2026:

Updated the docker wrapper to v22.

https://github.com/BOINC/boinc/wiki/Docker-wrapper-release-notes

What about forwarding the BOINC-Client-Proxy-Setting to the Podman/Docker-App ?

My BOINC-Clients are configured to use my separate running Squid, but the Podman/Docker-Tasks still say:

No local HTTP proxy found.

To use a proxy for the podman engine and/or pass a proxy setting into a container use the containers.conf file.
Extended information can be found running 'man containers.conf'.
In short:

[engine]
env = [
     "http_proxy=http://proxy_xy.example.com:3128",
     "https_proxy=http://proxy_xy.example.com:3128",
     "no_proxy=localhost,127.0.0.0/8,192.168.0.0/24"
   ]

In reply to computezrmle's message of 8 Jun 2026:
To use a proxy for the podman engine and/or pass a proxy setting into a container use the containers.conf file.
Extended information can be found running 'man containers.conf'.
In short:

[engine]
env = [
     "http_proxy=http://proxy_xy.example.com:3128",
     "https_proxy=http://proxy_xy.example.com:3128",
     "no_proxy=localhost,127.0.0.0/8,192.168.0.0/24"
   ]

[/quote]
I entered the command in WSL Boinc-Buda-Runner, but it doesn't know the man-command

In the Boinc-Buda-Runner, I have 4 containers.conf. Which one is the right one ?

\\wsl.localhost\boinc-buda-runner\etc\containers
\\wsl.localhost\boinc-buda-runner\mnt\wslg\distro\etc\containers
\\wsl.localhost\boinc-buda-runner\mnt\wslg\distro\usr\share\containers
\\wsl.localhost\boinc-buda-runner\usr\share\containers

But, editing these files can only be a temporair workaround, BOINC has to manage this by itself when the User has setup the Proxy-Info in the normal Client

Yeti