LHC@home

Theory tasks with the powheg-box generator are appearing again.
This time the init process name is run-Wj.exe. No idea how long that process will run before the 71000 events are processed.

Theory tasks with the powheg-box generator are appearing again.
This time the init process name is run-Wj.exe. No idea how long that process will run before the 71000 events are processed.

Uhm... I dunno whether it's the same thing... but maybe I remember something alike.

You're running Windows, I run native Theory on Linux. That said:
I remember that I've encountered some Theory tasks (I can't tell you they were powheg-box or not, sorry) which weren't updating the runRivet.log file even for one whole day, before beginning the processing step... and sometimes finishing it in a while. From what I can recollect, in the initial generating step they were writing generated events to files like

/var/lib/boinc/slots/slot_number/cernvm/shared/tmp/tmp.something_alphanumeric/run-main/pwgevents.lhe

Each event had its own section of X lines in this file (one task had maybe X=7, another maybe X=11...). Something like:

<event>
yada
yada
...
</event>

Knowing from runRivet.log that N events had to be generated, you could estimate the ultimate size of the pwgevents.lhe file: about X*N lines.
And if the file was Y lines long at a certain point, you knew that the events generating process was at (Y/(XN))*100%.

That to say:
a) an estimate of the generating process duration was possible;
b) even if runRivet.log hasn't been updated for many hours (even more than 24), that doesn't mean the task is stuck.

If I missed Your point completely, if I misunderstood Your message, please forgive me: it wasn't on purpose.

Bye.

That to say:
a) an estimate of the generating process duration was possible;
b) even if runRivet.log hasn't been updated for many hours (even more than 24), that doesn't mean the task is stuck.

If I missed Your point completely, if I misunderstood Your message, please forgive me: it wasn't on purpose.

Thanks for your explanation on the native tasks, but since I'm on Windows VBox, I have almost no access to the Virtual Machine except the provided ALT-F-keys and
the output of ALT-F2 (normally showing events processed) is not helpful during the init-phase of the powheg-box tasks.
Normally I shutdown the Theory's, but now I will let it run overnight.

Started yesterday:

[INFO] ===> [runRivet] Mon Aug 26 16:29:06 UTC 2024 [boinc pp w1j 7000 250 - powheg-box r3744 default 71000 393]

The run-Wj.exe init-process has run over 24 hours and until now no event (of the 71000 to do) has been processed.

2024-08-27 18:02:15 (7356): Status Report: Elapsed Time: '84000.613651'
2024-08-27 18:02:15 (7356): Status Report: CPU Time: '82988.140625'

https://lhcathome.cern.ch/lhcathome/result.php?resultid=413827128

The previous mentioned task with the powheg-box generator finished succesfully: https://lhcathome.cern.ch/lhcathome/result.php?resultid=413827128
The run-Wj.exe init-process ran 41 hours before the job started processing the 71000 events.
The event processing part lasted about 2 hours.

The previous mentioned task with the powheg-box generator finished succesfully: https://lhcathome.cern.ch/lhcathome/result.php?resultid=413827128

Glad to read it! :-)

Bye.