Container-native monitoring is key to the success of your DevOps initiatives
Want your DevOps initiatives to succeed? According to the State of the DevOps Report, you should probably look into container-native monitoring. Peter Arijs explains more.
The State of DevOps Report every year presents key findings on how DevOps practices lead to higher IT performance. The 2017 report was just released and there are some interested trends, in particular related to the impact of faster release cycles.
The report defines DevOps as an understood set of practices and cultural values that has been proven to help organizations improve their software release cycles, software quality, security, and the ability to get rapid product feedback. This higher performance delivers improved business outcomes, as organizations are able to better meet their goals. This is of particular importance in a time of IT transformation, where organizations strive to accelerate their digital initiatives.
Loosely coupled architectures and teams are the strongest predictor of continuous delivery, is one of the key findings of the report. Indeed, loosely coupled services that can be developed and released independently, allow organizations to deliver software faster and more incrementally. Containers and microservices are important supporting technologies for this.
The report also states that as organizations move to such architectures, there is still a difference in results, in particular for what they qualify as “high” and “low” performers. A summary of this is shown in the table below, but for more details, see the full report. When compared to the 2016 results, the difference between high and low performers narrowed for throughput (deployment frequency and change lead time), but widened for stability (mean time to recover and change failure rate). A lot of lower performing organizations are thus working hard to improve throughput, but without building sufficient quality in the process.
Failure, and especially failing fast and recovering fast, is partly inherent to the DevOps approach, but as the results show, care has to be taken such that the pendulum does not swing in the wrong direction. In an environment with very fast release cycles, frequent changes to the production environment are made. New containers are added, services are updated, etc. In such an environment, it is of vital importance to retain visibility on the impact on these frequent changes—both on technical and business performance—and proactively act on stability issues, as many high performers do.
Particularly in containerized environments—which are often the underlying architecture—visibility can be challenging. Traditional monitoring tools and strategies are not always well adapted to these dynamic environments, and can result in blind spots and slower time to detect and recover issues. Full visibility into container environments has unique requirements, as I discussed previously. In addition to technical performance metrics, your monitoring strategy should also keep track of the impact on your key business metrics, which are often unique to your line of business.
As your organization strives towards accelerating IT transformation initiatives and faster release cycles, make sure to build sufficient quality in the process. A monitoring strategy that is well adapted to these dynamic and containerized environments is an important part this puzzle, albeit certainly not the only one. Many high performers are already using adapted tools to understand the impact of these fast release cycles and to proactively act on stability and performance issues.