After upgrading a CA API Gateway Virtual Appliance from 9.2 to 9.4 with Expedited Appliance Upgrade.
(https://docops.ca.com/ca-api-gateway/9-4/en/install-configure-upgrade/upgrade-the-gateway/expedited-appliance-upgrade)
The new API Gateway 9.4 doesn't start up properly and repeats restarting with the following SEVERE level message in ssg_0_0.log.
2019-07-10T09:53:39.278+0900 SEVERE 1 com.hazelcast.instance.Node: [192.168.1.123]:8777 [gateway] [3.10.2] Could not join cluster in 300000 ms. Shutting down now!
How can this problem be resolved?
Release : 9.4
Component : API GATEWAY
In this environment, the original API Gateway 9.2 host was up and running in the same network segment.
The new API Gateway 9.4 recognized the original 9.2 Gateway as the MASTER node of a cluster and tried to join to the cluster but they weren't compatible each other.
The problem was resolved by clearing the cluster_info table in the ssg database.
Please run the following command In the privileged shell.
# mysql -e "delete from ssg.cluster_info;"
After restarting the ssg service, the API Gateway will start up properly.