Oracle does provide statistics that allow us to provide users with an approximation percentage of how much more time is needed for the current LogMiner query.

This may be useful in situations where users may think the connector is stuck or not processing, but can see that it is progressing and approximate the remaining time needed.

So the proposed metrics might include things like this

• CurrentFetchBytesRead
• CurrentFetchBytesRemaining
• CurrentFetchBytesTotal
• CurrentFetchTimeRemainingInMilliseconds
• CurrentFetchTimeElapsedInMilliseconds

Now because these need to be run adjacent to the existing LogMiner query, a secondary background thread would be needed to query and populate these results to the JMX metrics beans rather than relying on doing this in the main capture thread.