Indexing by SchedulerName instead of SchedulerId in SchedulerRepostory makes it difficult to connect to multiple nodes in a cluster. I have to maintain my own repository and make sure I call SchedulerRepostory.Remove after connecting to each node (GetScheduler on a StdSchedulerFactory with proxy enabled), otherwise the same RemoteScheduler instance is returned no matter what specific node I'm trying to reach.

I need this to support interruption in my Quartz cluster. Each node configures their quartz.scheduler.instanceId with their exported remoting host:port, so I can figure out how to reach each running job using the QRTZ_FIRED_TRIGGERS:INSTANCE_NAME field. This will also allow me to query nodes for job execution progress/results and fetch scheduler metadata.