Upgrade Procedures

concept

Step-by-step procedures are provided for upgrading an Operational Insights cluster, online or offline.

This section provides step-by-step instructions for upgrading a multi-node cluster from Enterprise Analytics to Operational Insights. Each procedure suits a particular situation. To choose between them, see Upgrade Procedure Selection.

Upgrading Multi-Node Clusters

Upgrade an Offline Cluster. This is the simplest procedure, and works from every Enterprise Analytics release, but it needs cluster downtime. Each node follows a per-node procedure: Offline Upgrade from Enterprise Analytics to Operational Insights for nodes installed from packages, or Offline Upgrade of a Docker Cluster from Enterprise Analytics to Operational Insights for nodes that run in Docker containers.
Upgrade an Online Cluster. The cluster continues to serve requests throughout. Three procedures are provided:
- Upgrade a Reduced-Capacity, Online Cluster. If no spare machine is available, the cluster can be upgraded with its own machines, provided that it can serve its workload with one node fewer during the upgrade.
- Upgrade a Full-Capacity, Online Cluster. If a spare machine is available, the cluster stays at full capacity throughout.
- Upgrade an Online Docker Cluster, Full Capacity. For a cluster that runs in Docker containers.

Upgrading Single-Node Clusters

Single-node clusters are often used for development, and are not supported for production. For their limitations, see Single-Node and Two-Node Clusters.

A single-node cluster has no other node to take over its work, so upgrade it offline: follow Upgrade an Offline Cluster. Because there is only one node, you do not need to disable and re-enable auto-failover.