WAN Optimization with Application Performance Management
Business software is usually provided centrally to multiple sites through a wide-area network – increasingly from a cloud. Voice-over-IP, multimedia applications, videoconferencing and mobile solutions have significantly increased the data traffic carried by these networks. Existing WANs are unable to cope: they lack the bandwidth to accommodate higher volumes of traffic, and as a result, latencies are increasing. The typical response to performance degradation is to add more bandwidth.
But there is a smarter alternative: using Application Performance Management (APM) to analyze data traffic and make better use of bandwidth. This approach reduces duplicate transfers of identical data, harnesses smart data compression methods, and prevents delays caused by unnecessary transfer protocol overhead. As a result, WAN speeds can be increased up to 50-fold with existing bandwidth. To guarantee high quality in terms of the user experience, APM continuously analyzes the ICT infrastructure, issues alerts when application response time thresholds are exceeded, and provides timely visibility into issues, faults and bottlenecks.
The significant rise in WAN traffic is primarily a result of the widespread adoption of IP technology, video conferencing, unified communications and collaboration, and cloud computing. But conventional applications can also cause network congestion and slowdowns. Plus, the transfer and application protocols that allow computers to communicate with each other and with the servers in data centers are not always optimized for WANs. Application Performance Management eliminates unnecessary overheads in WAN traffic, reducing it essentially to the payload data.
Reducing latency
This improvement on its own is extremely effective in networks linking data centers with sites in other countries – because in this scenario latencies grow considerably while IT performance decreases. This can be illustrated with an example from an international logistics business: Originally, its Asian sites could not use mission-critical applications provided from a data center in Europe. Network analysis by Application Performance Management experts, up to and including the application layer, identified ways of achieving significant improvements. Today, these applications are running smoothly across Europe and Asia, eliminating the need to build a data center in Asia – and delivering significant savings in terms of capital investment, ICT and human resources.
Data compression
Application Performance Management can also cut data traffic through intelligent, real-time compression. This is made possible by mathematical techniques and high-performance hardware. Similar methods are employed in video communications. Compression reduces the size of data packets before transmission, enabling more payload to be transferred with existing bandwidths.
Monitoring of applications, networks and hardware
Continuous monitoring – a service component of Application Performance Management – sheds light on the actual behavior of applications in everyday use. Monitoring works on three levels: the applications, the data-center and client hardware, and the networks themselves (including components such as routers and switches). A range of methods is employed. Voice Health monitors VoIP performance and quality, while Traffic Health looks at bandwidth consumption by applications. Net Health gives businesses visibility into LAN/WAN bandwidth usage and how the capacity of devices within the infrastructure is being utilized. Application Health focuses on the application layer and user experience. All monitoring data is available both in real time and as historical information. This accelerates and simplifies troubleshooting and enables businesses to make the right investment and expansion decisions quickly and efficiently.
The stability and performance of all business applications depends to a high degree not only on data centre services, but also on the quality of local area networks (LANs).
Challenges such as expanding data traffic, virtualization and cloud computing are ramping up the pressure on data center LANs – making a corresponding network infrastructure, tailored to this specific scenario, a must.
