This version of GitHub Enterprise was discontinued on 2021-09-23. No patch releases will be made, even for critical security issues. For better performance, improved security, and new features, upgrade to the latest version of GitHub Enterprise. For help with the upgrade, contact GitHub Enterprise support.

Initiating a failover to your replica cluster

If your GitHub Enterprise Server cluster fails, you can fail over to the passive replica .

About failover to your replica cluster

In the event of a failure at your primary datacenter, you can fail over to the replica nodes in the secondary datacenter if you configure a passive replica node for each node in your active cluster.

The time required to fail over depends on how long it takes to manually promote the replica cluster and redirect traffic.

Promoting a replica cluster does not automatically set up replication for the existing cluster. After promoting a replica cluster, you can reconfigure replication from the new active cluster. For more information, see "Configuring high availability for a cluster."

Prerequisites

To fail over to passive replica nodes, you must have configured high availability for your cluster. For more information, see "Configuring high availability for a cluster."

Initiating a failover to your replica cluster

  1. SSH into any passive node in the secondary datacenter for your cluster. For more information, see "Accessing the administrative shell (SSH)."

  2. Initialize the failover to the secondary cluster and configure it to act as the active nodes.

    ghe-cluster-failover
  3. After the configuration run finishes, GitHub Enterprise Server displays the following message.

    Finished cluster configuration
  4. Update the DNS record to point to the IP address of the load balancer for your passive cluster. Traffic is directed to the replica after the TTL period elapses.

After GitHub Enterprise Server returns you to the prompt and your DNS updates have propagated, you've finished failing over. Users can access GitHub Enterprise Server using the usual hostname for your cluster.