• Application performance management (APM) is becoming the most important application lifecycle management component for PaaS offerings as DevOps seeks better insight into app performance
• A paradigm shift is taking place as large organizations move operations to the cloud, steering the focus from infrastructure capacity planning to the application layer
APM is being reinvented in a new cloud era to provide developers and operations with a window into advanced application performance. So much is at stake to ensure applications perform effectively that APM technology is being baked into the fabric of the application at its earliest test cycle phase to help developers combat problems before they commit to code.
The emergence of containers and microservices architectures with on-demand scaling and hybrid cloud scenarios has impacted the need for new development processes for app performance, monitoring, and logging. APM providers are forming entirely new internal teams to innovate around this new era, namely by building a cloud-native offering which gets fed back into its broader traditional APM technology. The solutions will become an important component of cloud offerings. Consider Cisco’s recent high-profile $3.7 billion acquisition of AppDynamics, a leading APM provider which ensures Cisco’s success with end-to-end analytics, and brings together endpoint, network, and application logic. Major PaaS vendors IBM, Red Hat, SAP, Red Hat, and others are in the process of establishing strategies to better support their modern development architectures and platform services with the technology born of pure plays including Dynatrace and Datadog.
APM provides insight for a better understanding at runtime of which resources the application is using, why it’s running slowly, and when to turn off resources when they’re not being accessed. This once straightforward process under a monolithic single instant server app deployment has transformed into a suite of APIs and discreetly running microservices. Integrating APM during the beginning app design/development stage ensures a better understanding of the user experience in relation to the app performance, so if the app starts degrading deep in the bowels of an API connection, the APM can provide a rendering of multiple independently running processors to help diagnose why an app is failing.
Borrowing from technologies which were born of the big web scale companies like Google, Twitter, and Yahoo to address the complexities of applying monitoring tools for distributed apps involving scalability, there’s a strong movement behind open source software (OSS) technologies. The idea is to unify various approaches towards monitoring, service meshing, and distributed tracing framework through projects including Hawkular, Prometheus, Istio, Linkerd, and OpenTracing.
Please see this report for more on how this market is filling gaps in the ALM requirements of the DevOps process, including a recap of leading players in the space.