Bashbot federation

[mathew-fleisch]

Multiple bashbot instances (on separate servers) should be able to use the same trigger/config, but only execute a command in one instance.

Example: bashbot is deployed on Jenkins master/worker nodes to get status, clear space and other routine maintenance tasks from one bashbot config, but have the configurability to run bashbot commands in one or more random nodes, one or more specific nodes, or all nodes.

[mathew-fleisch]

I have been thinking about this (in the context of kubernetes) and I think there should actually always only be a single replica of bashbot. Instead, bashbot could include N number of (optional) executors/workers that when a command is detected from the master pod, is processed on one of these workers rather than on the master pod itself.