AWS load blanacer setup

[Chungwei]

I followed the documentation to setup the ant media server on AWS with auto scaling enabled, however, I use the t2.micro type as the instance type, so that it is easy for me to test the server, in terms of auto scaling.

But it seems like the AWS load balancer doesn't perform a good traffic routing, although the number of instance did scale, but it keep route the traffic to one instance, which caused the particular instance goes down, and all the viewer that used the instance got disconnect.

Is there anyway to solve this problem?

[BURAK KEKE]

Hi,

Did you try it for which protocol in publishing side (WebRTC or RTMP) and in playing (WebRTC or HLS)?

Also, t2.micro doesn't have sufficient resources for AMS? Maybe this causes the problem?

[Chungwei]

Hi,

Both sides are using WebRTC.

And yeah, t2.micro is definitely not sufficient enough for AMS, but I use for dev environment just to easily mimic the auto-scaling scenario. So I'm not sure if this will happen to t2.medium or greater.

My current solution is to write a simple node server in different port to perform the CPU usage check, as the AWS load balancer health check, not sure if this is a good solution, but at least it performs a much better traffic routing.