Parallel query and number of connections.
Hi there, we are running postgresql 10.5 on a centos 7 server.
We're seeing multiple connections (in pg_stat_activity) from our
application with the same query , same user, same application_name, same
query_start etc.
We are 100% sure the query is duplicated and not referring to multiple
queries issued.
The only difference is obviosly the pid and client_addr, that is populated
in any of the connections but the first one.
I was wondering if this can be related to the parallel queries of
postgresql. Is it supposed to spawn multiple "connections" ?
Thank you in advance,
Alessandro.
On 11/21/18 9:45 AM, Alessandro Aste wrote:
Hi there, we are running postgresql 10.5 on a centos 7 server.
We're seeing multiple connections (in pg_stat_activity) from our
application with the same query , same user, same application_name, same
query_start etc.We are 100% sure the query is duplicated and not referring to multiple
queries issued.The only difference is obviosly the pid and client_addr, that is
populated in any of the connections but the first one.
client_addr being different would seem to indicate that this is
something other then parallel query.
I was wondering if this can be related to the parallel queries of
postgresql. Is it supposed to spawn multiple "connections" ?Thank you in advance,
Alessandro.
--
Adrian Klaver
adrian.klaver@aklaver.com