[pgpool-general: 8077] ERROR: unable to read data from DB node 0

[Nikolay Mastilo (HM)]

Hello,

We use Pgpool as load balancer and connection pool in our application 
with three database nodes cluster (Patroni as failover manager).

Previously, we received such messages in the log when starting Pgpool:

2021-11-17 13:58:40: pid 19: LOG: find_primary_node_repeatedly: waiting 
for finding a primary node
2021-11-17 13:58:40: pid 19: ERROR: unable to read data from DB node 0
2021-11-17 13:58:40: pid 19: DETAIL: EOF encountered with backend
2021-11-17 13:58:40: pid 19: LOG: find_primary_node: 
make_persistent_db_connection_noerror failed on node 0
2021-11-17 13:58:40: pid 19: ERROR: unable to read data from DB node 1
2021-11-17 13:58:40: pid 19: DETAIL: EOF encountered with backend

These messages appeared during the first 5 minutes of the Pgpool 
starting, then stopped appearing.

When we set the parameter BACKEND_FLAG0 = 
ALWAYS_PRIMARY|DISALLOW_TO_FAILOVER , such messages no longer appear on 
startup.

Could you explain what happened and why such messages stopped after 
setting the specified parameter value.

-- 

Regards,
Nikolay

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pgpool.net/pipermail/pgpool-general/attachments/20220411/726b7f06/attachment-0001.htm>