How could I find the estimate cost for a 5000 concurrent meeting? Each group will consist around 10-30 people.
Assuming you are talking about self hosting: This many concurrent users is going to get tricky. It will require in-depth Linux server infrastructure knowledge as well as excellent understanding of how the difference Jitis Meet services interact with each other.
You will want load balanced OCTO enabled geographically distributed “CPU-Optimized Droplets on Digital Ocean” or Amazon AWS auto-scale groups for your multiple Jitsi Video Bridges (JVBs). See: https://calculator.aws/
@corby That’s quiet complicated tbh I setup jitsi with docker swarm and it’s trouble me enough just to get it working. To load balance the video bridge, probably gonna need server expert for this. The unfortunate part is, we don’t have luxury to hire any infrastructure expert just yet. I’ll go with Horizontal + Vertical scaling.
I was testing the bandwidth usage on digital ocean and after running it and monitor the bandwidth with 10 hours. It’s turn out that the usage is around 250-700mb per hour (depend on quality) but my limited testing, it has a constant 1 Mbps bitrate which is a low quality option. I’m happy with the result so far but it’s only 30 concurrent user. I will need a way to stress test 1000+ concurrent users.