Implementing a Multi-Server Chassis

Deprecated

With Storage 10.0, Swarm is single-process only. This information applies only to Storage 9.x and earlier.

This section describes how to enable the multi-server functionality in Swarm to create two or more independent nodes in a single physical chassis.

Platform Server

If you use Platform Server, skip this section: your hardware is set up.

Environments that include high-density hardware with multiple cores can benefit from the multi-server functionality available in the Swarm 3.0 and later releases. This feature allows you to create multiple independent Swarm nodes in a single physical chassis with a subset of drives assigned to each node.

Requirements

For multi-server implementations, each node requires at least:

Two CPU cores
Two hard drives
2 GB RAM

Note

If your node does not boot from a CSN, you must assign a static IP address to the node.

See Configuration Settings for Multi-Server.

Effect of multi-server

After you set up a multi-server configuration, shutting down or restarting one node impacts all of the nodes in the multi-server chassis. For example, if you set up a four-process multi-server chassis and restart a single process (a single node), the entire four-node chassis restarts. This is true for both SNMP shutdown/restart and shutting down or restarting the chassis using the Swarm Admin Console because all services in the multi-server chassis share the same resources.

Monitoring and administration

The Swarm Admin Console displays each physical chassis at the same level as a subcluster.

Note

In a multi-server implementation, use the Shutdown and Restart SNMP commands or Admin Console actions with care. Restarting the chassis or shutting down a single process within it (which appears as equivalent to a node on the Swarm Admin Console) restarts or shuts down all associated processes on the same chassis because they share system resources.

The Retire action impacts only the volumes for the specified process.

To retire all volumes within a chassis, initiate retire actions for all associated server processes within the chassis.