SIGSEGV Error when PREDICT

[giadapignataro]

Hi,

I am trying to create MODEL_DATA column with Predict Mode using DDF inside pipeline.
I get this error and DDF dies, but memory is not full.

• 13:49:39 - ClassMS | Putting column MODEL_DATA in /local/work/pignataro/A399_A401_lofar/Download/mss/a399-a401/A399-A401/mss-extract/calibration/TC00.MS
• 13:49:41 - AsyncProcessPool | worker 'io00' killed by signal SIGSEGV
• 13:49:41 - AsyncProcessPool | 1 worker(s) have died. Initiating shutdown.
• 13:49:49 - DDFacet | A worker process has died on us unexpectedly. This probably indicates a bug:
• 13:49:49 - DDFacet | the original underlying error may be reported in the log [possibly far] above.
• 13:49:49 - DDFacet | There was a problem after 11m48.8s; if you think this is a bug please open an issue,
• 13:49:49 - DDFacet | quote your version of DDFacet and attach your logfile.
• 13:49:49 - DDFacet | You are using DDFacet revision: 0.6.0.0
• 13:49:49 - DDFacet | Your logfile is available here: /local/work/pignataro/A399_A401_lofar/Download/mss/a399-a401/A399-A401/img/wideDD-c00.log

Using a lofar node with 128CPU and 512GB RAM.
Attaching the complete log file

wideDD-c00.log

[bennahugo]

Very hard to tell what is going on here. I know there was an issue in the LOFAR beams package some time ago. Can you please pull bhugo/ddfacet:0.7.2 from Dockerhub and retry - this should have a newer LOFARBeams package.

@cyriltasse any ideas?