I have "full-load-and-cdc" migrations tasks set up to track changes made in MySQL database and replicate changes to AWS S3. The cluster is InnoDB and I connect to one of the replicas. Sometimes, it happens that instance has to be detached from cluster and then added back as clone. In this situation all replication tasks fail with errors like this: "Could not find first log file name in binary log index file".

So, it looks like BinLog file with a specific name doesn't exist anymore. What could I do in a situation like this and be sure I continue replication from last checkpoint?