On 6 September 2017 at 01:52, milist ujang <ujang.milist@xxxxxxxxx> wrote:
-- Hi all,due to space issue and high volume transaction, some wal segments removed from pg_xlog on bdr environment.
What, you deleted them?
I had played streams and goldengate (oracle product) , that at capture side we can move forward to certain sequence (archivedlog/redolog - wal segment in postgres).So, is it possible to move forward to read recent wal segment in bdr environment? (assume data reconciliation will be done manually).
BDR can, see bdr.skip_changes_upto .
But PostgreSQL's logical decoding requires a contiguous WAL stream to maintain a valid catalog_xmin and restart_lsn, so it'll still fail to advance. So your replication from that node is broken, and you have to part the node then rejoin. You'll need to manually recover any diverged data.
Don't go in and randomly delete things in the postgres data directory, or things will break.
The BDR manual warns of the importance of disk space monitoring...
