Development for LHC@home

I suggest to copy the first line from .../slots/x/cernvm/shared/runRivet.log to stderr.txt.
This would be helpful to identify the scientific app in case of an error.

Example from the currently running task:

===> [runRivet] Wed Feb 13 09:09:33 UTC 2019 [boinc pp mb-inelastic 2360 - - pythia6 6.424 default 100000 16]

I suggest to copy the first line from .../slots/x/cernvm/shared/runRivet.log to stderr.txt.
This would be helpful to identify the scientific app in case of an error.

Example from the currently running task:

===> [runRivet] Wed Feb 13 09:09:33 UTC 2019 [boinc pp mb-inelastic 2360 - - pythia6 6.424 default 100000 16]

This should be done in the latest version but I have to see a successful task first to confirm.

I suggest to copy the first line from .../slots/x/cernvm/shared/runRivet.log to stderr.txt.
This would be helpful to identify the scientific app in case of an error.

Example from the currently running task:

===> [runRivet] Wed Feb 13 09:09:33 UTC 2019 [boinc pp mb-inelastic 2360 - - pythia6 6.424 default 100000 16]

This should be done in the latest version but I have to see a successful task first to confirm.

It didn't work. I have pushed out a new version (cranky-0.0.17) which I hope fixes this.

Just got a couple of these:

08:23:36 (8122): wrapper (7.7.26015): starting
08:23:36 (8122): wrapper: running ../../projects/lhcathomedev.cern.ch_lhcathome-dev/cranky-0.0.17 ()
../../projects/lhcathomedev.cern.ch_lhcathome-dev/cranky-0.0.17: line 26: xmllint: command not found
14:23:36 2019-02-14: cranky-0.0.17: [INFO] Detected  App
14:23:36 2019-02-14: cranky-0.0.17: [INFO] Checking CVMFS.
../../projects/lhcathomedev.cern.ch_lhcathome-dev/cranky-0.0.17: line 44: [@]: bad substitution
14:23:36 2019-02-14: cranky-0.0.17: [INFO] Checking runc.
14:23:55 2019-02-14: cranky-0.0.17: [INFO] Creating the filesystem.
14:23:55 2019-02-14: cranky-0.0.17: [INFO] Using /cvmfs/cernvm-prod.cern.ch/cvm3
14:23:55 2019-02-14: cranky-0.0.17: [INFO] Updating config.json.
14:23:55 2019-02-14: cranky-0.0.17: [INFO] Running Container 'runc'.
../../projects/lhcathomedev.cern.ch_lhcathome-dev/cranky-0.0.17: line 26: xmllint: command not found
/cvmfs/grid.cern.ch/vc/containers/runc: "run" requires exactly 1 argument(s)
14:23:55 2019-02-14: cranky-0.0.17: [ERROR] Container 'runc' failed.
08:23:55 (8122): cranky exited; CPU time 0.079445
08:23:55 (8122): app exit status: 0xce
08:23:55 (8122): called boinc_finish(195)

Just got a couple of these:

../../projects/lhcathomedev.cern.ch_lhcathome-dev/cranky-0.0.17: line 26: xmllint: command not found

It may help to install xmllint.
See:
https://lhcathomedev.cern.ch/lhcathome-dev/forum_thread.php?id=446&postid=5893

It may help to install xmllint.

Ubuntu 18.10 says "Unable to locate package xmllint"
What next?

It may help to install xmllint.

Ubuntu 18.10 says "Unable to locate package xmllint"
What next?

I will try to solve this issue later today. Was hoping that this would be installed by default.

It may help to install xmllint.

Ubuntu 18.10 says "Unable to locate package xmllint"
What next?

Installed it with

apt-get install libxml2-utils

It may help to install xmllint.

Ubuntu 18.10 says "Unable to locate package xmllint"
What next?

It's included in "libxml2-utils" or "bash-completion".

See:
https://packages.ubuntu.com/search?lang=en&suite=xenial&arch=any&searchon=contents&keywords=xmllint

I suggest to copy the first line from .../slots/x/cernvm/shared/runRivet.log to stderr.txt.
This would be helpful to identify the scientific app in case of an error.

Example from the currently running task:

===> [runRivet] Wed Feb 13 09:09:33 UTC 2019 [boinc pp mb-inelastic 2360 - - pythia6 6.424 default 100000 16]

This should be done in the latest version but I have to see a successful task first to confirm.

It didn't work. I have pushed out a new version (cranky-0.0.17) which I hope fixes this.

Your hope was honored: https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=2752565

[snip]
19:30:43 2019-02-14: cranky-0.0.17: [INFO] Preparing output.
===> [runRivet] Thu Feb 14 17:24:25 UTC 2019 [boinc pp jets 7000 25,-,100 - herwig7 7.0.3 UE-MMHT 100000 16]
20:30:44 (30726): cranky exited; CPU time 7584.431972
20:30:44 (30726): called boinc_finish(0)
[/snip]

I see time differences. It probably means that the job info is not added directly after the start of the job,
what could cause 'no job-info' when a task starves early.

I see time differences. It probably means that the job info is not added directly after the start of the job,

Correct. I can change this.

what could cause 'no job-info' when a task starves early.

I don’t understand what you mean.

what could cause 'no job-info' when a task starves early.

I don’t understand what you mean.

When you change adding the job-info to std output directly after the start it's present when a job runs into an error.

I see time differences. It probably means that the job info is not added directly after the start of the job,

Correct. I can change this.

Done

what could cause 'no job-info' when a task starves early.

I don’t understand what you mean.

When you change adding the job-info to std output directly after the start it's present when a job runs into an error.

It is not so straight forward as the output goes to the runRivet.log and this only exists some time after the job starts. I can revisit this later when we need to debug these kinds of errors.