On Tue, Mar 26, 2019 at 09:50:37AM -0600, Rob Sargent wrote:
> No, sorry I should have said that up front. We’re simple folk.
What is the WAL position (LSN) where Postgres is writing to and what
is the set of WAL segments in pg_wal (or pg_xlog if that's a server
older than 10)? Please double-check the configuration value of
wal_keep_segments, and as mentioned upthread, could you make sure that
you have no replication slots active? This can be done simply by
querying pg_replication_slots. Please note as well that checkpoints
are server-wide, so there is no point to run them on all databases.
Only one command will be effective for all databases.
--
Michael