Best Order to restart Pega Platform 8.1.2 on 15 servers
We have 15 servers with jboss EAP and Pega installed. Every server has different NodeType defined.
We have tried to stop all server and then start all server with no order and some of the processes such Cassandra or Kafka do not start properly.
But if we start one by one, sometimes they start correctly but other times some nodes are in a JOINING_FAILED status like RTGD nodes.
Is there any kind of dependency between them?
Is there any dependencies, for example, to join RTGD nodes to the cluster must be at least one DDS node in the cluster or simmilar?
What is the best way to make a complete restart of the Platform?
***Moderator Edit-Vidyaranjan: Updated SR details***
During start up ,all the required background process should get processed / executed/called/ started before end user can access the system for better performance hence web node is the last one in the order while stating the nodes and the reverse logic applies while shutting down the nodes.