VMware, Inc.
Role management of compute nodes in distributed clusters

Abstract:

In one example, a distributed cluster may include compute nodes having a master node and a replica node, an in-memory data grid formed from memory associated with the compute nodes, a first high availability agent running on the replica node, and a second high availability agent running on the master node. The first high availability agent may determine a failure of the master node by accessing data in the in-memory data grid and designate a role of the replica node as a new master node to perform cluster management tasks of the master node. The second high availability agent may determine that the new master node is available in the distributed cluster by accessing the data in the in-memory data grid when the master node is restored after the failure and demote a role of the master node to a new replica node.