One of the primary reasons why many IT operations teams have tried to limit usage of Docker containers in a production environment is lack of visibility. Most IT organizations today rely on tooling that doesn’t support containers to monitor and manage their IT environment. To solve that issue, providers of monitoring tools such as AppDynamics, a unit of Cisco Systems, developed services such as Microservices IQ to provide visibility into containers.
AppDynamics today announced it is extending the reach and scope of Microservices IQ by tightening the integration of the service with the Docker Monitoring utility. In addition, AppDynamics also announced the ability to generate heat maps to make it easier to visually identify individual containers or clusters of containers that are performing below normal.
The previous version of Microservics IQ already provided the ability to logically track any number of microservices associated with a specific application or business transaction, in addition to providing a thread contention analyzer that identifies where a thread is blocked all the way down to a specific line of code.
Matt Chotin, senior director of developer initiatives for AppDynamics, says by integrating Microservices IQ, an IT organization can now drill down into metrics for both containers and the host server they run on via the same pane of glass. That information can then be correlated to a specific business transaction to provide the context an IT organization needs to initiate the appropriate response.
Given the inherent ephemeral nature of containers, Chotin says a heat map will prove essential. Most microservices span multiple containers that are replaced regularly. As such, the condition of any specific microservice is subject to change at any given moment, especially when a new container in the environment starts to consume more memory or disk space than the previous container it replaced.
The challenge IT organizations face today is that most end users are unforgiving when it comes to application performance, Chotin says. A recent App Attention Index study published by AppDyamics found that 8 in 10 users have deleted apps because they have not performed correctly. The study also finds more than half have deleted an application or abandoned a website after just one attempt due to performance issues.
To make it easier to identify those issues, AppDynamics added a Tier Metric Correlator feature to Microservices iQ that visualizes load imbalances and performance anomalies across all the nodes in a tier. Chotin says issues that would have taken hours to diagnose using multiple dashboards spanning multiple products can now be identified in as little as a few seconds using the same application performance management (APM) tool they currently rely on to troubleshoot legacy applications.
Over time, data generated by AppDynamics will be shared with a Tetration analytics application based on the Apache Spark in-memory computing framework, Cisco has made a cornerstone of its efforts to inject machine- and deep-learning algorithms into DevOps. In the meantime, IT operations teams should become more comfortable deploying containers on-premises or in a public cloud computing environment as APM tools become more tightly integrated with container metadata. In fact, armed with that information, DevOps teams should one day be able to automatically remediate any container issue almost as fast as it arises.