Kafkacdc errors for wrong GTID positions are not clear

The error that is logged is something like this:

error  : Failed to read replicated event: 1236, Error: connecting slave requested to start from GTID 0-123-123, which is not in the master's binlog. Since the master's binlog contains GTIDs with higher sequence numbers, it probably means that the slave has diverged due to executing extra erroneous transactions

Since this is always due to the kafkacdc connecting with the wrong GTID, the error could state that the stored GTID state should be discarded. Better yet, this could be automatic so that kafkacdc automatically goes into gtid=oldest mode when this happens and then continues to replicate from the oldest GTID position. Some sanity checks would be needed to make sure that the GTID position that gtid=oldest results in is actually newer than the original GTID.