Managing Performance for Today’s Cloud-Based Applications

Never before has the performance of web-based applications had such a sizeable impact on business and revenues. Despite multi-megabyte, last-mile connections, many end-users are frustrated by the slow speeds they experience in practice. The result? Lower application adoption and lost revenues for businesses.

A key fact that’s often overlooked is that as page load times increase, so does the rate at which users abandon pages. A recent study illustrates how a poor end-user experience has a direct impact on revenue and ROI – a page load time of just six seconds results in a page abandonment rate of about 20 percent. If the page in question is a shopping cart checkout, then that business is losing 20 percent of its customers – and revenues – simply because of page load time.

With so much at stake, businesses must carefully manage the performance – the speed and availability – of all their critical applications, including those that are based in the cloud. But besides the cloud, there are many potential trouble spots standing between a business and their application end-users which may degrade performance. These include one’s own data center (which has become significantly more complex through multiple tiers and technologies like virtualization), third-party services (capabilities like RSS feeds and plug-ins can dramatically slow page load speeds), service provider backbones (which are often oversubscribed), peering points (which may deliver poor connectivity between networks), all the way to the end-user’s “last mile,” or their ISP connection and browser. The advent of the cloud adds yet another element of complexity and potential failure to this long application delivery chain.

This new world requires a different approach to managing application performance. This article will explore how IT can better manage this complexity and ensure superior performance for their applications living in the cloud, and how the only way to effectively do this is to proactively measure performance from the only perspective that matters – that of the end user.

The Cloud Presents Complex Challenges

Cloud computing represents a compelling new way to achieve business agility, flexibility and time-to-value. But it also introduces serious risks when it comes to managing the performance of applications. Why?

First, the cloud is opaque, meaning you lose precious visibility into application components and infrastructure when you move to the cloud.

Second, as a shared infrastructure, the cloud’s performance for all applications tends to suffer when any single tenant experiences a sudden spike in demand. Contrary to popular belief, cloud processing and bandwidth is not without limits and a traffic spike for one tenant often means less infrastructure availability for other tenants, at least temporarily. In other words, “infinite elasticity” doesn’t mean “instant elasticity.” These architectural limitations of the cloud should concern application owners, as the cloud’s impact on web applications is growing fast. According to IDC, the worldwide market for public IT cloud services grew at a blistering pace in 2010, reaching $21.5 billion. IDC forecasts public IT cloud services spending to reach $72.9 billion worldwide in 2015 — a compound annual growth rate (CAGR) of 27.6 percent — and a tripling of public IT cloud spending.