[RTADPA BW Tests] replace some 'problematic' files from hlt1 bw input
Fix PR/AllenInMoore_hlt1_bandwidth
nightly test due to some issues accessing certain files. These are now removed and new files have been added in their place to guarantee >100k events so we can do the nightly test with the precision we require.
Tested + should be merged with lhcb-datapkg/PRConfig!370 (merged)
Merge request reports
Activity
added RTA label
assigned to @lugrazet
added PR/AllenInMoore_hlt1_bandwidth label
- Resolved by Luke Grazette
There's no software shifter/maintainer in the DB for this week, but I see only a single ci-test currently running so I'll take the liberty of starting a restricted test here myself with lhcb-datapkg/PRConfig!370 (merged).
Potentially linked to the failure seen in the nightly AllenInMoore test. -> Locally found and then fixed errors running AllenInMoore Hlt1 test, but not technically the same one reported in the nightlies. Hence adding the
PR/AllenInMoore_hlt1_bandwidth
label, I want to double check this fixes the issue on the actual runners./ci-test lhcb-datapkg/PRConfig!370 (merged) --platforms x86_64_v3-centos7-gcc12+detdesc-opt+g
Edited by Luke Grazette
added ci-test-triggered label
- [2024-01-09 10:33] Validation started with lhcb-master-mr#10043
- [2024-01-11 15:51] Validation started with lhcb-master-mr#10078
- [2024-01-12 15:44] Validation started with lhcb-master-mr#10098
Edited by Software for LHCbadded hlt1-throughput-increased label
added 1 commit
- 933ba098 - rm 00204940_00007281_1.digi, as it's breaking the test (on the runners, not locally?)
Hi @msaur, may I have a small ci-test here when possible, including lhcb-datapkg/PRConfig!370 (merged) only on the platform
x86_64_v3-centos7-gcc12+detdesc-opt+g
.I want to test the nightly test runners as unfortunately the error we are getting I can't reproduce locally
but I think might be resolved by these changes.- Resolved by Luke Grazette
/ci-test --platforms=x86_64_v3-centos7-gcc12+detdesc-opt+g lhcb-datapkg/PRConfig!370 (merged)
added hlt1-throughput-decreased label and removed hlt1-throughput-increased label
requested review from @rjhunter
mentioned in merge request lhcb-datapkg/PRConfig!370 (merged)
mentioned in issue lhcb-core/LHCbPR2HD#8 (closed)
- Resolved by Sebastien Ponce
Hi @rjhunter, I'm making some fairly small and inoffensive changes to remove some broken files (here) and to more accurately use IOHelper depending on input file type (in lhcb-datapkg/PRConfig!370 (merged)) to repair the AllenInMoore hlt1 bandwidth test. (see !2834 (comment 7472859) for fixed test and comments).
Opened lhcb-core/LHCbPR2HD#8 (closed) to track the handler Quality-of-life improvement so that in the meantime we can prioritise getting these MRs into master and fixing the nightly.
Let me know if you have any comments before we progress
Edited by Luke Grazette