Redundancy and failover

Running Elemental Live events and Elemental Statmux MPTSes in Conductor Live lets you implement several resiliency features. These features help reduce outages in your workflows.

To clarify the terms:

• Resiliency refers to the ability to continue when errors occur.

• Redundancy refers to duplication of hardware or software components to protect against single points of failure. Therefore, redundancy is one way to achieve resiliency.

Conductor Live offers resiliency solutions in all areas of the workflow:

• Node redundancy for Conductor Live nodes, Elemental Live nodes, and Elemental Statmux nodes. This redundancy protects against failure of an entire node.

  Node redundancy is the foundation for some resiliency options within the worker nodes. For example, within Elemental Live, there are some resiliency options that only work with specific types of redundancy. Keep this fact in mind when planning node redundancy.

  When you first deploy your Conductor Live cluster, you should plan redundancy for the nodes in the cluster.

  After your initial deployment, you should review your node redundancy design when you add more nodes or when your workflows change dramatically.

  For more information about planning node redundancy, see Setup: Planning resiliency in a cluster.

• Resiliency features in different types of workflows:

  • Encoding workflows: workflows that involve only Elemental Live.

  • MPTS workflows: workflows that involve Elemental Live and Elemental Statmux.

  As part of the design procedure for a workflow, decide which resiliency features you want to implement. An encoding workflow and in an MPTS workflow have slightly different resiliency options.

  For more information about resiliency features, see Resiliency features in Elemental Statmux.