We have 2 PGP Universals in a cluster running 2.9.0b397 as appliances on two physical servers. To update DNS settings we broke up the cluster, made the DNS update (the only change), then rejoined them to the cluster per admin guide. Now we cannot login to the web interface of the secondary node with the same admin credential we've always used. Credential still works on the primary node, but is rejected as incorrect on the secondary. We were able to login to the secondary node with the same credential right after we broke up the cluster.
Clustering log on primary shows "authenticated cluster member <...>" after the cluster was recreated. The clustering log on the SSH console shows many ping messages between primary and secondary, but I cannot tell if replication is going on or not.
"pgpsysconf --restart pgprep" and "pgprepctl" do not work on either node, showing up as command not supported. Power cycling did not help.
As there's only the one admin credential, is there a way to reset it via the secondary node's SSH console?