To pick this back up again, I think the version of LINC I was running should be commit 0fd88967
alex (9142586e) at 27 Mar 14:25
add CI job
alex (c8b55147) at 27 Mar 14:24
undo commit
alex (ee34d7fe) at 27 Mar 14:22
add CI job
To add some context perhaps: if not specified (like in Roland's command), my runners always clone the master branch of LINC.
Hi Alex,
I have used git pull. I pulled the latest copy of LINC on the 20th of March, and that's the version that then encountered the error.
Cheers,
Roland
Hi Roland,
this means you have not updated LINC via git pull
?
If not then I am not certain which version you were running. Please contact Frits.
Cheers, Alex
Hi Alex,
Sure! I run LINC using Frits' runner: https://github.com/tikk3r/flocs/blob/upgrade-to-f38/runners/run_LINC_target.sh
bash run_LINC_target.sh -r $(pwd) -d Data -c ../3C295/L2022677_LINC_calibrator/results_LINC_calibrator/cal_solutions.h5 -s ${MYSINGULARITYDIR}/flocs_v5.0.0_sandybridge_sandybridge.sif -b /cosma5 -e"--selfcal=True --selfcal_strategy=HBA --target_skymodel=target.skymodel --make_structure_plot=False"
I will re-run the pipeline using the two options you specified.
Cheers,
Roland
Could you please elaborate a bit more how you run the pipeline? Unfortunately, I can not reproduce your issue.
Furthermore, please also try to re-run the pipeline with --debug
and --leave-tmpdir
and check the results of the chunking step (or post the logs here).
Alex
Hi Alex,
Yes, I am running LINC within Frits' 5.0.0 singularity container.
Cheers, Roland
Do you actually run LINC inside or outside the container?
Hi Alex,
Thanks for your response! I have indeed selected selfcal, but the weird part is that I just pulled a fresh LINC copy right before the run last week. The fix related to issue #56 (closed) apparently does not work for me.
Cheers, Roland
Dear @RolandTimmerman,
do you have used selfcal in your HBA run? We had the same issue reported here: #56 (closed), which was fixed 2 weeks ago.
Cheers, Alex
Frits Sweijen (c922b8ea) at 22 Mar 13:36
Add LOTSS as skymodel source option
I'm running into the issue that my LINC target seems to crash during the save_logfiles step:
[1;30mINFO[0m [workflow ] starting step save_logfiles
[1;30mINFO[0m [step save_logfiles] start
[1;30mINFO[0m [workflow ] starting step save_results
[1;30mINFO[0m [step save_results] start
[1;30mINFO[0m [workflow ] starting step save_inspection
[1;30mINFO[0m [step save_inspection] start
[1;30mINFO[0m [job save_results] /.../tmp.FctKGFGVmu/tmpdir_LINC_target/00k9bj6v$ bash \
collect_files.sh
results
cp: missing destination file operand after 'results'
Try 'cp --help' for more information.
[1;30mWARNING[0m [33m[job save_results] exited with status: 1[0m
[1;30mWARNING[0m [33m[job save_results] completed permanentFail[0m
(Data path abbreviated for ease of read)
Full log attached to this issue. I just pulled again LINC two days ago, so I should be on the most recent version.
Please let me know if you need any other information!
Cheers,
Roland
This MR does not aim for fixing #57, but should allow the users to run LINC again. It might cause issues in the summary step. So this branch is EXPERIMENTAL!
alex (9b3052ae) at 20 Mar 11:05
temporary dirty fix
Hi,
My LINC target run crashed while working on the structure function. Here is the excerpt from the job log:
[1;30mWARNING[0m [33m[job structure_function] exited with status: 1[0m
[1;30mERROR[0m [31m[job structure_function] Job error:
("Error collecting output for parameter 'output_plot': LINC/steps/LoSoTo.Structure.cwl:60:7: Did not find output file with glob pattern: ['3C277.3ComaA_structure.png'].", {})[0m
[1;30mWARNING[0m [33m[job structure_function] completed permanentFail[0m
Also, I have attached the relevant cal_solutions.h5-losoto_err.log.
cal_solutions.h5-losoto_err.log
Any and all advice is welcome! Please let me know if any more information would be helpful.
Thanks,
Roland
Hi Alex,
Thanks! I am indeed now running with the structure function disabled. Hopefully that will finish just fine. If not, I think I may open a new issue as then there will likely be a different underlying problem.
Cheers,
Roland