Compared to even a few years ago, managing the application ecosystem has become an indecipherable puzzle. The widespread adoption of virtualization and cloud technologies is saving companies lots of money and creating true efficiencies. But it’s also making the task of managing applications more complicated. Applications or components of them are now agile and constantly moved and re-provisioned as demand and conditions change. And applications themselves are more complex, sprawling multiple tiers, with code updated more frequently and often linking to other applications and/or third-party content sources that enterprises have limited control over. And last but not least, the proliferation of mobile devices and Web apps is creating more uncertainty around users’ activities. As a result we have less visibility and control into the apps, which makes supporting them more difficult. This post will look at a technique that we see customers asking for more and more in the face of all this application complexity: real user monitoring.
Given all this complexity in the application ecosystem, you’d think users would cut IT some slack on application response times. Of course not!
Whether they’re employees or customers, end-users today demand lightning-fast applications and the ability to work anywhere, on any device. Not long ago people would wait over 5 seconds for an application to respond. Now they get frustrated after about two seconds.
As a result, real-user monitoring (RUM) has become critical. Real-user monitoring gives you the performance intelligence and visibility you need for today’s application eco-system. It allows you to get a handle on all these cloud, mobile, and third party challenges. Whether a user accesses an application from a desktop at home, laptop in a hotel room, or from a smart phone in a café, real-user monitoring is a key component of today’s performance management solutions.
However, RUM can be implemented multiple ways, and enterprises should take time to consider the right approach for their specific environment. Here are four approaches to consider when planning a real-user monitoring strategy.
1. Build monitoring code into the application: Initially, application developers built code into the application to collect the monitoring data. This meant they were altering the application, which can be disruptive. Some organizations prefer this approach, however, because they may have only a few pages to instrument. Or their development teams may be designing applications with performance monitoring and management in mind.
2. Application Delivery Controllers (ADC): Using an ADC is a very scalable method because it eliminates the need to alter your application. Instead, ADCs inject the instrumentation code into the application stream without requiring changes to the application code. In addition to being very scalable, this approach supports dynamic and quickly changing applications. However, it does require a delivery controller that supports content injection.
3. Browser-based monitoring: If you’re using SaaS applications, deploying an ADC is not possible because you don’t have control over the application’s infrastructure or code. Instead, you can install a downloadable browser plug-in tool that tracks all the Web traffic to and from the user’s device. This is ideal for applications that you don’t host or own.
4. Monitoring the native mobile app: Some enterprise applications (and many more consumer applications) are best delivered through a native mobile app. However, this presents a challenge for performance monitoring, because the application is executing on the user’s device, and often not based on HTTP or a browser interface. This makes browser-based monitoring and ADC code injection impossible. A great solution here is to use open source software development kits (SDKs). An SDK can collect information on the mobile OS, connectivity speed and user performance and enable your development teams to quickly instrument their mobile apps with minimal effort.
Closing the feedback loop:
IT doesn’t exist for IT’s sake. At the end of the day, the IT team needs to deliver applications and data to end-users in a manner that supports their engagement or productivity. Real-user monitoring provides the necessary intelligence to make critical decisions about application infrastructure and identify problems when they occur.
For an example of Riverbed’s approach to end-user experience monitoring, you might want to read a recently published solution brief from Enterprise Strategy Group. The paper examines how Riverbed’s latest release of OPNET AppInternals Xpert bridges the gap between operations and development teams by combining end-user experience, transaction tracing, and application component monitoring for deep, end-to-end visibility into application. Download the paper here.
Riverbed. WAN optimization for your network: Application acceleration, WAN bandwidth optimization, and IT consolidation. Riverbed is the IT performance company. WAN optimization solutions from Riverbed liberate businesses from common IT constraints by increasing application performance, enabling consolidation, and providing enterprise-wide network and application visibility – all while eliminating the need to increase bandwidth, storage or servers. Thousands of companies trust Riverbed to deliver greater productivity and cost savings by making their IT infrastructure faster, less expensive and more responsive. Riverbed solutions are also available as managed services through select providers.