When I first started writing about Application Performance Management (APM), the technology was largely focused on root-cause application diagnostics and business transaction management (BTM). Companies like the now-defunct Optier were riding high based on their ability to track transactions through multiple layers of complex infrastructure. Many APM tools were used primarily in the development phase (or after a problem had already occurred) because the server overhead caused by monitoring agents would actually slow the performance of the very applications they were monitoring.
APM tools were costly, and the complexity of configuration, implementation, integration and ongoing support was overwhelming for large and small businesses alike. As a result, organizations were reactive, rather than proactive, and it took time to diagnose and fix problems, impacting productivity, customer satisfaction and, ultimately, the bottom line.
Today’s APM market offers a broad range of choices for customers (beyond those mentioned below), but choosing the right vendor and the right tool or set of tools can be difficult and can result in purchasing features that aren't needed. When evaluating APM tools, customers should consider the following:
- The problems you are trying to solve
- Existing investments in both APM solutions and infrastructure
- Application environments (distributed, mainframe, mobile, cloud, etc.)
- The applications you want to monitor and the type of applications they are (mobile, Web, transaction-oriented)
- On-premise or SaaS delivery model
- Level of expertise of IT staff and who within the organization will be using the tools
- The type of monitoring information that is most important in your business (user, transaction, etc.)
- Your implementation time frame
Not every tool will be the right fit for your company, so organizations should spend time evaluating exactly what they need before making any decisions.
Ease of use features and new delivery models
Against this backdrop, new companies (ExtraHop, New Relic, Splunk, AppDynamics) entered the market with solutions that are quicker to deploy, easier to manage and agent-less or with “lightweight” agents for low server overhead. These are, in general, less costly, and, in some cases, vendors offer software-as-a-service (SaaS) delivery models to drive down the barriers to entry even further.
As an example, the ExtraHop solution (available in a licensed model or as SaaS) is a completely passive self-contained appliance that monitors transactions on the network to gain a comprehensive, real-time view into the performance of all tiers of the application infrastructure, including load balancers, web servers, application servers, databases, networked storage systems and even third-party cloud computing services. According to ExtraHop, the platform can be deployed and begin analyzing application transactions in under an hour.
By the same token, traditional APM vendors such as CA Technologies (formerly CA Wily), with CA APM, and Compuware (now Dynatrace) added new features and SaaS delivery models that reduce cost, improve ease of use and reduce overhead. These solutions support a “DevOps” approach to application development where both development and operations work collaboratively to shorten the application release cycle. By employing APM throughout the product life cycle — in both development and operations — application quality is improved and customers get problems solved more quickly and new features sooner. [Disclosure: CA Technologies is a client of Clabby Analytics.]
Vendors such as INETCO continue to focus on monitoring real-time transactions based on their industry focus in payment processing and banking. INETCO Insight captures and correlates multi-protocol transactions across multiple banking sub-transactions, while operating independently of the underlying application or switch being monitored and without requiring agents, easing implementation and integration and reducing overhead.
Many vendors include analytics capabilities that track, monitor and correlate multiple metrics to predict an outage or potential performance issue before it occurs. Netuitive was a pioneer in this market, offering "self-learning" analytics that automate performance analysis with behavior learning technology, correlation discovery and anomaly detection. Other APM vendors (including those mentioned above) have also added analytics, dashboard visualization and "views" for a variety of roles within the organization.
Today one of the primary areas of focus for APM is user experience, based on skyrocketing growth in user-facing Web and mobile applications. Businesses across a range of industries, including banking, telecommunications, retail and healthcare, are trying to engage with customers using innovative Web and mobile applications to improve customer service, provide a seamless digital experience and sell more products.
Today’s users are less likely to be brand loyal, and customer churn can be minimized by offering superior application performance on multiple devices. While most vendors offer some type of user experience monitoring, Aternity Inc. specializes in user experience monitoring with an approach that takes every device across physical, virtual and mobile environments and turns it into a self-monitoring platform that is user–experience-aware. By doing this, the solution provides accurate insight into what end users see from whatever device they are using across an entire portfolio of applications. [Disclosure: Aternity Inc. is a client of Clabby Analytics.]
Dynatrace has recently spun-off from Compuware, enabling the company to focus solely on APM and launching the Dynatrace brand. With a commitment to digital performance management, omni-channel customer experience monitoring, support for a broad range of DevOps use cases, and purpose-built analytics, the company is reporting 31% growth in new customers.
Support for new application environments
New application environments such as cloud and mobile have also added complexity to application monitoring. Many of today’s composite applications span multiple layers of infrastructure including distributed environments, mobile, and mainframe — on-premise and off-premise — and linking these silos for a single unified look at application performance is an important feature for many enterprise customers. A number of vendors including IBM, Nastel and CA Technologies offer a suite of integrated tools for monitoring across all these platforms. [Disclosure: IBM is a client of Clabby Analytics.]
For example, CA APM’s Mobile APM provides a 360-degree view of mobile to mainframe application behavior. When used with CA Mobile App Analytics (MAA), administrators can monitor transactions from mobile to middleware to mainframe, providing a complete view of application behavior for easy identification of potential performance problems. Another example is Nastel’s AutoPilot that spans distributed, middleware and mainframe environments performing real-time monitoring with a particular focus on software and appliance-based middleware, including WebSphere MQ, TIBCO EMS and RV, Solace, WebSphere DataPower and WebSphere Message Broker.
This is not meant to be an exhaustive list, but rather some things to think about as you look at various options. Also bear in mind that many vendors today offer free downloads and a free 30-day trial period to try the product on a few applications and/or a few users. This can be a great way to determine if the solution works for your business and meets your objectives.
This article is published as part of the IDG Contributor Network. Want to Join?