If you are managing a wide area network (WAN), chances are your network will be supporting the delivery of a variety of applications to your user population. Traditionally, your principal concern was network performance, especially as it related to bandwidth, latency, throughput, and security. However, in today's increasingly dispersed and virtualized operating environments, you need to understand how well your network is performing against other criteria, like application delivery.
To ensure you know how well your WAN is delivering applications, you need real-time situational awareness of network performance across all parts of the network -- meaning, you need greater network visibility. Through greater network-based monitoring, you will achieve greater application visibility.
When optimizing your WAN -- which impacts application delivery -- you need information or visibility into network performance. WAN optimization is not simply about reducing bandwidth utilization. You also need visibility into the WAN infrastructure so you can analyze whether your resources are being used in the best way.
What network visibility parameters do you need for application visibility?
WAN optimization can dramatically improve the user experience, but you can further improve performance by proactively managing it. To manage the user experience, you need information on how the WAN is performing. Such parameters for measuring WAN performance include:
- packet counts;
- bandwidth available and in use;
- data center performance data;
- remote site performance data; and
- information on transmission paths being used.
In order to effectively manage network performance and achieve end-to-end visibility of the network and applications, enterprises using WAN optimization technology need to obtain performance data across all parts of the network -- not just between the two connection points of a WAN optimizer.
Figure 1 depicts how an optimized network typically uses appliances to perform optimization, like data compression, on both ends of a connection. However, the optimized data being transmitted may not present a true picture of the network's performance. This is because optimized data being transmitted is usually not identical to the original or delivered data. You need something that analyzes WAN performance data generated across the entire network.
Figure 1: Application visibility of a WAN optimization appliance
Typically, data at either end of an optimized network is not optimized prior to entering an accelerated WAN link. So you must be able to view what's happening to data across the network -- for example (as depicted in Figure 2), at the data center, where it is not optimized; the network, where it is optimized; and at the remote site, where it is not optimized. Once network performance data has been captured, your WAN optimization management solution should also be able to measure a number of variables, such as:
- response times;
- network latency;
- device availability and utilization;
- network channel availability and utilization;
- network protocol and device errors; and
- data flow and response times.
Figure 2: Application visibility of a WAN optimization management system
WAN traffic visibility of all network elements is needed in order to gain application visibility; this data can then be analyzed and used for making network management decisions.
What happens to all the performance data you can see and subsequently obtain? Ideally, you will have an automated tool that can be programmed to analyze the data gleaned across the WAN, providing you with data that compares network performance against specific metrics and/or service level agreements (SLAs) you establish. In addition to the usual daily network performance reports, your automated tool should also be able to compare the data against pre-defined performance thresholds and red flag situations that may be indicative of impending problems.
Continue reading the second part of this article to find WAN application visibility solutions.