New approach to ensure quality user experience for rich media applications: use routers, switches and ‘smart’ endpoints to monitor the real traffic flow
User satisfaction is the ultimate metric for performance evaluation of rich media services. The only way to be sure of delivering highest quality of experience is by actually measuring quality of experience. Therefore, it is not surprising that most IT managers identify application performance as their single most critical problem.
Many traditional performance monitoring solutions are delivered as appliances. Although an increasing number of solutions are deployed on virtual machines or in some cases cloud offerings are available, there are nevertheless ongoing administrative overhead. Monitoring is best done pervasively unless you already know in advance where all the problems are going to be. Deploying these traditional monitoring solutions typically only provide partial view as it is not possible to deploy probes everywhere.
With IOS Performance Monitor, first made available in IOS 15.1(3)T, we introduced a new innovative approach to performance monitoring. A Cisco network element is embedded with native monitoring capabilities. It can collect packet loss, jitter, delay and response time information for performance evaluation of data, voice and video services. This approach effectively turns every network element as probe providing the most extensive visibility because all the flows must go through the network. This approach eliminates the need for deploying flow data collection solutions external to the network devices which would lead to increased costs for data collection and operation expense. Now, network managers can continue to leverage their Cisco IP network assets, resulting in substantial savings.
In addition to measuring conditions at the network elements, this solution works in conjunction with a medianet endpoint embedded with the Media Services Interface. This ‘smart’ endpoint has the same monitoring capability as network elements effectively acting as passive probe for the network. Upon detection of issues, this ‘smart’ endpoint can trigger mediatrace diagnostics to collect pertinent information while the problem is occurring. This dynamic collection mechanism makes the solution extremely scalable because it collects the right information at the right time. Traditional performance monitoring solutions have not been able to bridge the network and the application to provide extensive visibility, enable end-to-end troubleshooting, minimize data collection costs and eliminate the need for probes .