Scaling Jitsi VideoBridge on an AWS Server Quickly for Spike in Traffic

Hi Jitsi Community,

Thanks for all the hard work put into the project.

We have integrated and host Jitsi on our own company servers, which are on AWS and I have set up Autoscaling.

Our software products allow virtual events to take place and so we will often have a very low amount of traffic and then could have 3 or 4 events being hosted concurrently all starting at the same time. We have set a maximum limit of 50 participants. This will cause a sharp increase in usage of our servers.

We can could have one large server running but this would be very expensive as we have a lot of time when there would be no traffic on the site particularly to begin with.

I am not a DevOps man, but we thought if we put a low threshold on the CPU of the AWS autoscaling we could almost dedicate a server to an event but the issue we would have then is that if all the events started at the same time all the videobridges of each conference would still start on the same server as the capacity of that server will not have been reached. On top of that, I read it takes around 5 minutes for a server to initialise

Do you have any suggestions on how we can handle this?