[pgpool-general: 9046] Running pcp_detach_node on a backend set to DISALLOW_TO_FAILOVER

[Ian van der Linde]

Good day

We are currently running pgPool-II 4.4.2, and would like to be able to detach a node from an external failover manager. In order words, we would prefer to switch off pgPool's built in automatic failover, and instead handle it using other scripts (in this case using EDB Enterprise Failover Manager). 

Due to this situation, we have set all the backend flags to DISALLOW_TO_FAILOVER, with the expectation that we can still manually set a backend status to down using pcp_detach_node. When we try to do this, however, we see the following message:

pcp_detach_node -n 0 -h 127.0.0.1 -U pcpadmin
ERROR:  degenerate backend request for node_id: 0 from pid [394] is canceled because failover is disallowed on the node

Is there some way in which we can prevent any automatic failovers by pgPool but which allows us to run detach/attach/promote as we see fit?

Kind regards
Ian