Hi,
We're dealing with a 9.5.5 database with the symptom that, after a certain amount of time after restart, the startup process reaches a certain WAL segment, and stops. The startup process runs at 100% CPU, with no output from strace. There are no queries running on the secondary, so it's not a delay for queries. The WAL receiver continues to operate normally, and we can see sent_location / write_location / flush_location continue to move ahead in parallel, with replay_location stuck in that WAL segment.
Suggestions on further diagnosis?
--
-- Christophe Pettus
xof(at)thebuild(dot)com