We have an issue in PROD environment that we cannot reproduce in DEV environment,since we are not using a huge load of data in DEV, for which we would like to discuss with you guys.
See line 17384 which shows the metric being sent AFTER the job has completed at line 17378 in the log file.
It is suspected that this happens with large volumes of data when using multi-threading.
When analyzing the masking, we see that the masking was 100% completed, but it still generates an .err file that blocks the rest of the masking process.
When the job is restarted (which can take a long time because of the analysis of the cards to be masked), the result is indeed 0 masked lines because the job was successfully completed during the previous masking.