Product

The AppD Approach: One Size Agent Doesn’t Fit All

By | | 3 min read


Summary
Read more about how AppDynamics provides intelligent agents tailored to your specific platform, providing a highly precise level of monitoring and analytics.

When it comes to deploying agents for your APM solution, a one-size-fits-all approach may sound tempting, but like most one-size solutions, they simply can’t hold up to the countless scenarios encountered in today’s enterprises.  We take a different approach at AppDynamics, providing intelligent agents tailored to your specific platform that provide a highly precise level of monitoring and analytics, and do so with much lower overhead. We’ve spent the last 10 years creating and perfecting agents that can easily be deployed and managed at the scale required by today’s enterprise.

The core of any APM solution is the agent. It’s responsible for collecting and reporting all relevant data and metrics. But in order for an agent to provide the deepest level of insights, it needs to be extremely intelligent. It must understand the platform with which it’s interacting, which data points are the most relevant, and how to obtain them. And these requirements and interactions will change from one platform to the next.

For example, Java has many frameworks which expose different entry points, is multithreaded, and supports a wide variety of message queues and databases. In contrast, Node.js is single-threaded, utilizing an event-loop to handle callbacks, and is often integrated with a different ecosystem of databases and message queues (though of course there are overlaps). Our Java and Node.js agents understand how to interact with each of these platforms, taking advantage of their unique interfaces to gather the most critical metrics from your applications. It’s these highly-intelligent agents that allowed us to help DirectTV adopt an APM solution covering 6,000 physical and virtual servers running a mix of Windows, Linux, and HP-UX, as well as applications built using Java, .NET, PHP, and Node.js.

Another crucial consideration when deploying APM agents is overhead, as it’s critical an agent not add costs by consuming too many resources and in turn be disruptive to production applications. This becomes even more crucial for enterprises where tens of thousands of agents might be deployed across thousands of servers. By utilizing an agent-level baseline, and continuing to monitor and update this baseline over time, AppDynamics’ agents are able to capture a small amount of standard data under normal circumstances, keeping the overhead extremely low. When a deviation is detected from what the agent considers to be normal, it dives deeper, capturing a detailed report of what is happening on the system at that moment to ensure your team will have everything required to troubleshoot and resolve the issue.

Today’s modern enterprises utilize every tool at their disposal to provide the best possible experience to their customers, and require an APM solution that works with all of them. AppDynamics understands this need and has built our agents to easily integrate into the latest deployment and management tools including Chef, Puppet, and Ansible, making it easy to deploy agents quickly. Consider Cisco (now our parent company but also one of our early customers) who was able to successfully deploy 15,000 AppDynamics agents across 2,000 physical servers and 5,000 virtual machines, serving more than 200 applications in 23 data centers – all in just one week!

We also know that security is extremely important, and that root access is something that SecOps teams tightly control. With this in mind, our agents are designed to perform their monitoring and information gathering without requiring root access.

Ready to find out the difference an intelligent APM agent can make for your business? Start a free trial today!

Rob is a Product Manager at AppDynamics, working with hundreds of customers over the last five years who use AppDynamics to gain visibility into their applications. Prior to AppDynamics, Rob worked in various roles at different technology companies like Akamai Technologies and Microsoft.