new art-test scripts added

New ART tests in InnerDetector/InDetExample/InDetSLHC_Example/test:

test_InclinedAlternative_17-06-00_single_electrons_100GeV.sh art_tests test_InclinedAlternative_17-06-00_single_muons_100GeV.sh art_tests test_InclinedAlternative_17-06-00_single_muons_10GeV.sh art_tests test_InclinedAlternative_17-06-00_single_muons_1GeV.sh art_tests

This merge request affects 1 package:

• InnerDetector/InDetExample/InDetSLHC_Example

Adding @goetz ,@sroe as watchers

added 21.9 InnerDetector review-pending-level-1 labels

CI Result SUCCESS (hash ce06c62f)

	Athena
externals
cmake
make
required tests
optional tests

Full details available on this CI monitor view
Athena: number of compilation errors 0, warnings 0
For experts only: Jenkins output [CI-MERGE-REQUEST 41403]

MR is all new files, escalating to L2. --L1

added review-pending-level-2 label and removed review-pending-level-1 label

pinging @ncalace ...

These are new scripts for ART tests

Personally I think this should be safe to merge, since these are new additions (so shouldn't break anything existing) and difficult to test fully outside of ART.

As far as I can see, the four scripts are identical except that the names of the variable are changed. To avoid duplicating the same code four times, this could be one script that has the variables set at the beginning depending on which case is wanted. If you need a separate script for each test, the heavy lifting could be in a common file and the separate scripts could call that script with the setting that change. ~Joe (L2)

added review-user-action-required label and removed review-pending-level-2 label

Yes, that's true. I just wanted to follow the path with the old scripts, but this should be definitely done in the way you suggest. Best, Marcin

Yes, this is a good point to reduce code clutter. Could we introduce these tests first and then do a clean-up in a second pass?

We could try it first of all in some of the older tests that these new scripts aim to replace, since it would be less of an issue if they don't run while we do the updates (I suspect that this may take a few iterations since it is difficult to fully reproduce how these tests will run in th ART framework when testing locally).

added review-pending-level-2 label and removed review-user-action-required label

Sounds reasonable to me.

@wolter, @nstyles, if you're not going to do it imminently then could you open a ticket lest the refactorign of these tests gets forgotten.

Tim (L2)

added review-approved label and removed review-pending-level-2 label

Hi, I have opened a ticket here: https://its.cern.ch/jira/projects/ATLIDTRKCP/issues/ATLIDTRKCP-211

merged

mentioned in commit 0c4ae096

This line is crashing in the ART at the moment. I assume its the missing space that is the issue, see lines from the joblog:

/cvmfs/atlas-nightlies.cern.ch/repo/sw/21.9/2019-11-22T2146/Athena/21.9.3/InstallArea/x86_64-slc6-gcc62-opt/src/InnerDetector/InDetExample/InDetSLHC_Example/test/test_InclinedAlternative_17-06-00_single_muons_100GeV.sh: line 178: --postExec: command not found
art-result: 127 reco
test_InclinedAlternative_17-06-00_single_muons_100GeV.sh: Reco_tf.py isn't working yet. Remove jobReport.json to prevent pilot declaring a failed job.

This might also be the case for the other tests. Could you please check and fix where necessary?

It would be good to fix this by tomorrow in order to get it in the next build => next ART run, so that we don't try and run incorrectly configured jobs again (which seems to cause the jobs to run very long).

The layoutTranslation should be removed... I'm surprised why this doesn't crash to be honest. This crash should have been visible locally too