Hi pgsql-admin email distribution list,
We are using the JDBC Driver bundled with the Debezium project. We are trying to reset the replication stream from a WAL log point in the past. This effort is in service of developing a disaster recovery plan, with the idea that a flushed LSN may have been in the Kafka Broker Pool but not consumed. The missing WAL files may still exist on disk, so it would be ideal if we could begin streaming from a given WAL log file.
To that end, we have configured the Kafka Connect daemon to begin streaming from a paritcular WAL log file.
Consider the following logs which show the creation of a new logical replication slot:
Although START_REPLICATION is called with the pg_xlog position we desire, streaming is picked up from the current WAL log position.
Is there a method to achieve our ends? Our alternative plan is to add a republish_count field to the the transaction, so that we can craft an SQL statement to update the rows from say, an hour ago to the present time.
Since WAL logs are not transaction logs (as I understand things) I believe the answer to this is no, but is it possible, from a given LSN, to go back X number of events in the WAL log and identify a transaction ID? If we had to go with plan B it seems like that could be a method for determining a starting point.
Thanks in advance for any assistance anyone can provide,
Joseph Hammerman
