I wanted to ask how does druid handles ingestion tasks where a few events may contain null as value of the timestamp column.
Basically, the timestamp column is present but a few events(say 10 out of 10000) contain null value for the timestamp column.
Does it skip those events and continue or the ingestion task fails??
It depends on your configuration. In particular, “reportParseExceptions” for most non-hadoop ingest methods, and “ignoreInvalidRows” for hadoop based ingestion.
Thanks, that worked for my batch ingestion tasks.
Is this config valid if I am ingesting events in realtime by using tranquility kafka.
The same should work for Tranquility Kafka.