Support Article
Servers fail with ClusterAvailabilityException
SA-62897
Summary
Servers fail with ClusterAvailabilityException.
Error Messages
Caused by: pega.decision.vbd.node.ClusterAvailabilityException: No cluster nodes available
at pega.decision.vbd.node.client.ServerNodeCache.getNext(ServerNodeCache.java:84)~[vbd-node-jar:?]
at pega.decision.vbd.node.client.DefaultClientClusterExecutor.execute(DefaultClientClusterExecutor.java:51) ~[vbd-node-jar:?]
at pega.decision.vbd.node.client.ClientEngine.insert(ClientEngine.java:64)~[vbd-node-jar:?]
at pega.decision.vbd.service.impl.VBDServiceImpl.insertFactRecords(VBDServiceImpl.java:158)~[vbd-impl-jar:?]
Steps to Reproduce
Unknown
Root Cause
The cluster crashes and the nodes stop communicating. Since the nodes do not receive signal from the other nodes (old node is terminated), they consider that they are broken and cast the members away.
Resolution
Here’s the explanation for the reported behavior:
According to the architecture, the existing node is replaced by a new node if any of the following occurs:
- High resource utilization and the node enters the hung state
- Application requires a longer time to open due to any activity
- Abnormalities in the server
A JAR file which requires a server restart to reflect is imported. Error message displays after the server restart since the nodes are unavailable to form a cluster. However, the server is stable after it receives the heartbeat message from a new node.
Published October 8, 2020
Have a question? Get answers now.
Visit the Collaboration Center to ask questions, engage in discussions, share ideas, and help others.