Monitoring the Health of Red5 Pro on AWS
You can configure AWS CloudWatch service to monitor your Red5 Pro installation.
Single Server Metrics
Single server instaces should be monitored for Availibility, CPU Usage, CPU Load and disk space at a minimum. If CPU Usage goes about 80% or CPU load approaches the number of vCPUs you have in your system then you should upgrade. Additionaly, disk space should be cleared or additional space added if go about 80% usage. Availability monitoring depends on your use case, but the following are suggested to be monitored via both the Public IP and the internal IP:
- Monitor web service port availability (port 5080, with SSL installed port 443)
- RTMP service port availabilty (port 1935)
Autoscale instances are ephemeral, but monitoring can be placed on the RDS and Elastic Load Balancers that service the cluster. If the RDS CPU usage goes above 60% you should consider upgrading. The following should be monitored via the Elastic Load Balancer Public IP as well as directly against the RDS Service:
- Monitor web service port availability (port 443)