[pgpool-general: 9517] Re: Problem with pentaho kettle job and pgpool

[Tatsuo Ishii]

> Hello.
> 
> We are currently working with a pgpool 4.5.2 (hotooriboshi) 3-nodes cluster, with postgresql 16.3, on a Debian 11.
> 
> We're facing a problem when trying to run a Pentaho kettle job. When we run the kettle on the pgpool VIP address and port (9999 in this case), it gets a point when trying to run a ktr file we get an error. In the pentaho log we get:
> 
> ERROR (version 9.4.0.0-343, build 0.0 from 2022-11-08 07.50.27 by buildguy) : Error occurred, process stopped
> Error occurred while trying to connect to the database
> Connection failed. Verify all connection parameters and confirm that the appropriate driver is installed.
> 
> Before reaching that point, the kettle runs other DDL sentences succesfully.
> The odd thing is that, if we direct the kettle to the primary postgresql node, "jumping over" pgpool, it Works correctly.
> I set the logging of pgpool in all three nodes in Debug5, but there are no clues on the logs.

One possibility is, num_init_children in pgpool.conf is too low and
max_connections in postgresql.conf is too high. If so, there's no
pgpool idle process and pgpool cannot accept connections from client.
Also can you share the output of pcp_proc_info to check whether "Wait
for connection" state pgpool process?

> The kettle runs on a Debian 12.
> pentaho pdi-ce-9.4.0.0-343
> openjdk version "17.0.13" 2024-10-15
> The driver is using is: postgresql-42.2.23.jar

Best regards,
--
Tatsuo Ishii
SRA OSS K.K.
English: http://www.sraoss.co.jp/index_en/
Japanese:http://www.sraoss.co.jp