Development for LHC@home

This version solves the issue from version 7.61 where the output file could not be found and so the result could not be uploaded to the server.

=================================================

Not related to this version. This behaviour was also there in v7.61.

When i run 10 Theory's on a 20 threaded machine (started with more than 1 minute interval), the jobs are sitting there for a very long time (even over 3 hours) without obvious doing nothing.

The position where the jobs are waiting is:

Building rivetvm ...
make: Entering directory '/scratch/rivetvm'
g++ rivetvm.cc -o rivetvm.exe -DNDEBUG -Wfatal-errors -Wl,-rpath /cvmfs/sft.cern.ch/lcg/releases/LCG_107a_ATLAS_3/MCGenerators/rivet/4.1.0/x86_64-el9-gcc13-opt/lib -Wl,-rpath /cvmfs/sft.cern.ch/lcg/releases/LCG_107a_ATLAS_3/hepmc3/3.3.0/x86_64-el9-gcc13-opt/lib `/cvmfs/sft.cern.ch/lcg/releases/LCG_107a_ATLAS_3/MCGenerators/rivet/4.1.0/x86_64-el9-gcc13-opt/bin/rivet-config --cppflags --ldflags --libs` -I/cvmfs/sft.cern.ch/lcg/releases/LCG_107a_ATLAS_3/highfive/2.10.1/x86_64-el9-gcc13-opt/include


Oh wait, after three hours something is happening in one slot. Some sign of life:

0 events processed
PYTHIA Warning in StringFragmentation::fragmentToJunction: bad convergence junction rest frame
PYTHIA Error in StringFragmentation::fragment: stuck in joining
PYTHIA Error in Pythia::next: hadronLevel failed; try again
PYTHIA Warning in JunctionSplitting::SplitJunPairs: parallel junction state not allowed.
PYTHIA Warning in JunctionSplitting::CheckColours: Not possible to split junctions; making new colours
PYTHIA Error in StringFragmentation::fragmentToJunction: caught in junction flavour loop
PYTHIA Warning in JunctionSplitting::CheckColours: Made a gluon colour singlet; redoing colours
PYTHIA Warning in SimpleSpaceShower::pT2nextQCD: weight above unity
PYTHIA Error in MiniStringFragmentation::fragment: no 1- or 2-body state found above mass threshold
PYTHIA Warning in MiniStringFragmentation::ministring2two: random axis needed to break tie

In reply to Crystal Pellet's message of 9 Aug 2025:

Not related to this version. This behaviour was also there in v7.61.

When i run 10 Theory's on a 20 threaded machine (started with more than 1 minute interval), the jobs are sitting there for a very long time (even over 3 hours) without obvious doing nothing.

After increasing the RAM for the Theory VMs to 1024MB this long inactivity behaviour disappeared. The machine I tested with has 64GB RAM.

Laurence, you may consider to change the line in the Theory_2025_08_08_dev.xml file from <memory_size_mb>630</memory_size_mb> into into <memory_size_mb>1024</memory_size_mb>

You also may consider to remove the line job_duration>864000</job_duration> from that xml-file. Nowadays we run always one single Theory job within the VM during one BOINC-task and not longer many jobs inside the same VM until the old 18-hours limit was passed.

For me this version with the little adjustments above is ready for production.

Many thanks. I have just released a new version with these two changes.

In reply to Laurence CERN's message of 13 Aug 2025:

Many thanks. I have just released a new version with these two changes.

The Theory_2025_08_13_dev.xml is the same as before.

In reply to Crystal Pellet's message of 13 Aug 2025:

In reply to Laurence CERN's message of 13 Aug 2025:

Many thanks. I have just released a new version with these two changes.

The Theory_2025_08_13_dev.xml is the same as before.

Sorry there was an issue with the release preparation script. It should be resolved now.

I have also made this change in production