Recently, a Data Connection ingest ran without making progress for 6 hours (versus the typical ingest time of 30 mins). This caused the source database server to grind to a halt, so we needed to kill the ingest process.
Is there a way to automatically abort an ingest if it runs for too long, so that if this job hangs again it won't cause the same issue?
Yes. You can set
maxAllowedDuration
on the jobspec in ISO-8601 format, e.g,maxAllowedDuration: PT10M
will automatically cancel the job after 10 minutes. Jobs will sometimes hang because of a database issue.