Repairing a Fabric Services System cluster

After the initial Fabric Services System installation, in the event of node failure (such as when a node is unreachable), you can repair the node cluster by removing the failed cluster and replacing it with another node without having to reinstall or redeploy the system.

To remove a failed node and replace it with another node, use the /root/bin/setup-k8s.sh utility in the deployer VM.

Repairing a Fabric Services System cluster involves the following tasks:
  1. Removing the failed node from the cluster
  2. Adding a replacement node to the cluster, which includes:
    • Updating the workernodes section of the input.json file and initiating the setup of the Fabric Services System environment
    • Running the setup-k8s.sh utility