CecSve
commented
3 months ago When I look at the error logs for that ingestion attempt&_a=(columns:!(_source),filters:!(('$state':(store:appState),meta:(alias:!n,disabled:!f,index:'439da4d0-290a-11ed-8155-a37cb1ead50e',key:level,negate:!f,params:(query:ERROR),type:phrase),query:(match_phrase:(level:ERROR)))),index:'439da4d0-290a-11ed-8155-a37cb1ead50e',interval:auto,query:(language:lucene,query:'datasetKey.keyword:%22747d8c01-0537-4979-ada7-dad9bfa0d59d%22%20AND%20attempt:%221%22'),sort:!('@timestamp',desc))) I can see the following error messages:
- GBIF ID is null, occurrenceId - 19fc0719-afd6-4432-92c1-f5d78ccc25a8 (the record referred to) - but the error message is thrown for multiple occurrenceIDs apparently (maybe all?).
- java.lang.IllegalStateException: Failed performing conversion on Caused by: java.io.IOException: Not a data file.
It seems to me that the IDs were missing in that attempt, but maybe you can explain @muttcg?
I published this dataset on June 10. It was picked up for ingestion the same day and everything seemed to be going smoothly.
Today I was alerted that data still had not shown up on the dataset page. Looking into the history in the registry, the process stopped at the VERBATIM_TO_INTERPRETED step. Any future attempts to recrawl the dataset were aborted because nothing had changed at the side of the IPT.
Today, I published a new version (with no changes) and forced a new ingestion. This ingestion went through with no problems and the data is available now on GBIF.
Any idea what the problem was on June 10, or possibly even with the dataset published at that time? I'm pretty sure there should not be any changes between the version on June 10 and the one I just published (but I don't know how I could verify this now).