I’ve faced yet another Greenplum Interconnect error with the following error:

"ERROR","58M01","Interconnect Error: Could not connect to seqserver.  (seg0 slice1 gpslave1:40000 pid=14231)","Connection refused (connect errno 111)"

After doing tons of checks on the gpslave1 server itself, nothing was found to be problematic. Ping shows reply, telnet to the ports shows active and no errors could be found.

Finally, on the master itself, I tried to ping the configured host of the master and it returned 127.0.0.1. That’s simply according to the /etc/hosts file.

127.0.0.1 localhost gpmaster
192.168.0.10 gpslave1
192.168.0.11 gpslave2
192.168.0.12 gpslave3

One more thing that’s left to be done; I changed the host gpmaster from 127.0.0.1 to the actual LAN IP and it works!

So, a small piece of note, all the nodes (including the master, should resolve to the IP the other nodes needs to access from. Simply assuming gpmaster would only need to listen to itself doesn’t work.