How to configure multiple hazelcast tcp-ip clusters in one yaml file?

Question

To configure one cluster in one yml file, I could do:

hazelcast:
  cluster-name: cluster1
  network: 
    join:
      multicast: 
        enabled: false
      tcp-ip:
        enabled: true
        member-list:
          - machineA
          - machineB

However, because of the way the yaml file is structured, I find it difficult to do a yaml file with 2 clusters. If I do it the following way, it would not work:

hazelcast:
  cluster-name: 
   - cluster1
       network: 
         join:
           multicast: 
             enabled: false
           tcp-ip:
             enabled: true
             member-list:
               - machineA
               - machineB
   - cluster2
       network: 
         join:
           multicast: 
             enabled: false
           tcp-ip:
             enabled: true
             member-list:
               - machineC
               - machineD

question from:https://stackoverflow.com/questions/65905503/how-to-configure-multiple-hazelcast-tcp-ip-clusters-in-one-yaml-file

Answer 1

You should define separate configurations per cluster. What you did is an invalid configuration. Say you have two different configurations: cluster1.yaml config and cluster2.yaml config. Then all the members of cluster1 should start with cluster1.yaml and the ones in cluster2 should use cluster2.yaml. If you need multiple clusters to be aware of each other - e.g being synchronized, consider using WAN Replication.

In that case the configurations will look like:

cluster1.yaml:

hazelcast:
  cluster-name: cluster1
  network:
    join:
      tcp-ip:
        enabled: true
        member-list:
          - A  # uses this config
          - B  # uses this config

cluster2.yaml:

hazelcast:
  cluster-name: cluster2
  network:
    join:
      tcp-ip:
        enabled: true
        member-list:
          - C  # uses this config
          - D  # uses this config