Scaling issues in OPENVIDU

How to implement?
1

Good morning, there is a project for education and today we have the following scenario at the peak:

Simultaneous sessions: 600
Format of each session: 1 publisher and 30 subscribers

Currently, with the support of the Openvidu team, I have the following configuration:

1 = Openvidu AWS server: c5.9xlarge (vCPU: 36 | Memory: 72 | storage: EBS only | Network bandwidth: 10 | EBS bandwidth: 9,500)

8 = Media node: m5.xlarge (vCPU 4 | Memory: 16 | storage: EBS only | Network bandwidth: Up to 10 | EBS bandwidth: up to 4,750)

Choose instance c5 as it is designed to use according to amazon’s description of this family to use:

  • Because Amazon EC2 C5 instances are the next generation of the Amazon EC2 compute-optimized instance family. Like C5 instances, the lowest price for the Amazon EC2 vCPU family and ideal for running advanced, heavy-duty workloads. This includes workloads such as high performance servers, high performance computing (HPC), batch processing, game serving, highly scalable multiplayer games, video coding, scientific modeling, distributed analysis and machine learning inference / deep learning.

Given the information above, which configuration would be ideal to meet the demand referred to above?

Best regards,

2

It is hard to say because it depends on users network quality, video resolution, etc.

As you have OpenVidu PRO, just monitor the usage and add more nodes on demand.

We are working on autoscaling feature right now, it will be available in 3/4 weeks.

Regards