Hi, We have a requirement to support 50–100 concurrent agent sessions. We are self-hosting the LiveKit worker and have enabled auto-scaling. However, since new machines can take a couple of minutes to spin up, we’re considering implementing custom queue logic where users wait until capacity becomes available. For this, I wanted to check: • Is there a way to determine how many workers are currently running? • Is there a way to monitor the number of active agent sessions? With this information, we could implement backend logic to place users in a queue and release them when capacity is available. Also, if there are better or recommended approaches for handling this scenario with LiveKit, I’d really appreciate your suggestions. Thanks in advance!

I think these are not exposed natively in livekit, but you can track externally whenever an agent job spins up

Thanks @flat-photographer-63184!