[pgpool-general: 4570] Online recovery: existing connections did not close in 90 sec.

[Jose Baez]

Hello!

I was doing online recovery with 2 PSQL nodes and Pgpool configured with
"replication_mode = on".

Every test I made recovering one PSQL node was succesful, but now I tried
to recover a bigger database (20 GB).

1st stage was made OK:

   - Put active psql node with WAL archiving.
   - Send pg_start_backup
   - Compress all PGDATA
   - Send pg_stop_backup
   - Remove WAL archiving
   - Transfer tar.gz to failed node (3 GB size)

pid 12936: 1st stage is done
pid 12936: starting 2nd stage

   - Now pgpool closes connections without waiting to finish queries (I
   configured that way).

pid 12728: pool_process_query: child connection forced to terminate due to
client_idle_limitis -1
...
...
pid 12817: pool_process_query: child connection forced to terminate due to
client_idle_limitis -1

And then:

ERROR: pid 12936: wait_connection_closed: existing connections did not
close in 90 sec.
ERROR: pid 12936: start_recovery: timeover for waiting connection closed


Why pgpool couldn't close a PSQL connection ?


Thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.sraoss.jp/pipermail/pgpool-general/attachments/20160323/abd57cb8/attachment.html>