Brief summary of the bug. What is it ? Where is it ?

The AbstractZeebeRecordFetcher.getZeebeRecordsForPrefixAndPartitionFrom method throws a RuntimeException in the case the index it tries to read from is missing.

This exception causes the import mediator to retry until an error retry timeout is hit.

This causes the import pipeline to get throttled in case some sub-entitites like incidents are not there.

AT:

• the case of a missing index is just treated as empty result, unblocking the import of other entitites