Autoscaling of Openvidu CE

OpenVidu v2 Issues developing apps
1

Hi everyone,

I was trying to autoscale the Openvidu CE version In AWS. I managed to isolate the “openvidu_app_1” and deployed it in the fargate and remaining containers (nginx, openvidu-server, kms and etc) are deployed on the EC2 and these are both are communicating using the internal ALB instead of localhost which is default behaviour. Now when i tried to use the application with One/multiple fargate containers and with one EC2 the application is working seamlessly, but the problem is when i increase the count of EC2 instances there are errors in the application.

error:
“token”:"wss://<domain>.com?sessionId=<sessionname>&token=tok_GcM9djkHaF1j9Up

wserror
wserror1920×216 158 KB

i tried the changes suggested in this link WebSocket connection to 'wss://mydomain/openvidu' failed: Error during WebSocket handshake: Unexpected response code: 400

but the error remained same, can anyone please help me resolving this issue.

Thankyou!!!

2

OpenVidu CE is not prepared to be deployed in a cluster composed of multiple OpenVidu CE.

Each OpenVidu CE has state, so you need to manage each OpenVidu CE instance individually.

3

hi @cruizba,
Thankyou for replying,
Is there any option in enterprise that can solve the above mentioned issue? i have gone through the docs related to pro and noticed presently pro has 2 limitations : Scalability - OpenVidu Docs so i was wondering that whether openvidu enterprise can handle this situation,

please suggest me any option that can resolve this issue.

Thankyou!!!

4

OpenVidu PRO and OpenVidu Enterprise provide different scalability options. They provide a preferred way to be deployed on AWS.

Other deployment ways are not supported.

Best regards