A network type is essentially a group of networks. The group should reflect the latency experienced by the users on the client network.
In general, the largest contributing factors to latency should be the geographic location and the bandwidth to the subnet. The management console includes default network types to indicate bandwidth, such as T1, however, the management console cannot assume where your networks are located relative to the monitoring devices that observes the application traffic. Therefore, it is important to customize network types to your organization’s layout because a big performance difference exists between a T1 that is 200 miles long and a T1 that is 1,500 miles long.
The number of network types required depends on the size of your organization. Use network types to group subnets that must talk across the same network paths and therefore experience the same latency. The key principal is that network types should be a group of subnets that share the same physical network resources and therefore experience the same latency due to distance, serialization, and queuing delay. Having an accurate network diagram helps greatly in building network types.
Note: By default, CA Application Delivery Analysis includes pre-configured VPN and Wireless network types for users coming from networks with highly variable latency. Performance thresholds for VPN and Wireless network types are pre-configured to one-half the sensitivity of other network types. As the administrator, you want the management console to create incidents that your team can resolve. In this case, it unlikely you can resolve latency issues with remote users accessing the network from the internet, for example, from a hotel in London or Singapore.
|
Copyright © 2014 CA Technologies.
All rights reserved.
|
|