Please describe why this is necessary.

While debugging issues in CDF, it is useful to know information about the query to pinpoint problems. I propose adding tracing statements to understand the metadata filtering in ChangeDataFeed.

Describe the functionality you are proposing.

Add tracing statement at the end of LogReplayScanner::into_scan_batches that logs the following:

• The size of the remove deletion vector hashset
• Whether cdc was found
• file path
• version
• timestamp
• Total number of rows processed
• Total number of rows selected

Additional context

No response