[pgpool-general: 6363] Re: pgpool performance issues

[Pierre Timmermans]

What do you mean by "My db has 500 max connections" ? If you have 500 concurrent connections, then you should set num_init_children to 500, because now you can have no more than 200 concurrent users (with pgpool a connection is released from the pgpool only if it the session disconnects from postgres)

Pierre 

    On Sunday, December 23, 2018, 9:26:17 AM GMT+1, Mariel Cherkassky <mariel.cherkassky at gmail.com> wrote:  
 
 Hi,I'm using pgpool and I'm suffering from very poor performance issues when using the pool. I have the next architecture  :1)Node A, contains the application,standby database and pgpool service.2)Node B contains the primary db and pgpool service that will be used only in failover.
I did the next 2 tests :-In the first test, the application on node A access the DB on node B directly, without connecting to the pool. I run one of our application`s major procedures and it took 40s.-In the second test, the application on node A access the pool on node A and it redirects the queries to node B. I run the same procedure and it took 3minutes.
My db has 500 max connections and the application has only 1 user and 1 db so I set the num_init_children to be 200 and pool_size to be 1.
Any idea how can I tune the pgpool ?
Thanks._______________________________________________
pgpool-general mailing list
pgpool-general at pgpool.net
http://www.pgpool.net/mailman/listinfo/pgpool-general
  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.sraoss.jp/pipermail/pgpool-general/attachments/20181223/b710d7a6/attachment.html>