Activate monitoring for BotsInterface's scheduler

Description

The BotsInterface and ExecutionInstance both have their own scheduler object, which isn't shared between them, even though they may point to the same underlying datastore. The ExecutionInstance scheduler had it's monitoring enabled if monitoring was configured, but the BotsInterface didn't. This was causing some job metrics to not be published, specifically execution metadata metrics as those ultimately get triggered by a job moving to Completed (which happens in the BotsInterface's scheduler).

To fix this, add support for enabling monitoring for the BotsInterface's scheduler, as well as allowing multiple schedulers for the same instance_name in Server. This is needed for registering monitoring queues and making sure metrics are published regardless of which instance of the scheduler they're coming from.