‘Docker’ and ‘Containerization’ are terms that seem to be as much of a buzz as IoT and Big Data today. While everyone loves talking about Docker and how it’s improving the agility of IT, the company acknowledges that adoption is still in it’s infancy for the most part. Interestingly, where Docker is seeing the most traction is within large enterprises, which Moogsoft defines as having 5000+ managed entities (the majority of our customers). According to a study conducted by StackEngine, 70% of enterprises are either evaluating Docker or using Docker in some form of testing and QA, development or production.
While development is using docker to do things like accelerate deployment times and enable flexible workload management, for example, it actually creates more complexity for ops to manage. IT incidents are now more frequent and obscure with the advent of containerization and enterprises now need to figure out how to manage and ultimately prevent these incidents from occurring and impacting service quality.
Here are 2 major challenges that Docker brings to IT operations:
1. Docker containers have a very short lifecycle
The primary attraction to Docker is the ability to rapidly deploy and move applications in a spontaneous fashion. Containers are rapidly spun up, down, and moved around as business needs dictate, meaning that the lifecycle is incredibly short. In fact, it’s been reported that the lifespan of a Docker container is 4x shorter than a Virtual Machine. The problem is that when an incident occurs, it’s incredibly difficult to keep up with the change and track exactly what was connected to what before, during, and after that incident. All stages likely contain bits of information that could be necessary to understand what really happened and identify the root cause vs. subsequent symptoms.
Troubleshooting IT incidents is like putting together a rather large puzzle and introducing containers makes troubleshooting like putting together that puzzle with moving pieces that constantly change shape.
2. More Docker containers means more event storms
We’ve established that Docker containers undergo rapid change. Any change to the composition of a container results in a myriad of performance metrics and event/alert notifications being generated. Furthermore, Docker containers are run on several hosts and the more hosts you have, the more events and alerts that will get fired when an incident occurs. As anyone that works in IT support at a large enterprise can tell you, event storming is already a huge problem. It’s a constant challenge to separate the signal from the noise and that noise and it’s only going to get louder with the adoption of containerization.
In response to these challenges, early adopters of Docker are turning to new tools that can help manage incidents more effectively. Moogsoft, for example, allows enterprises to ingest massive volumes of operational data from across their production stack, reduce the noise by 99%, and use machine learning to correlate the data to pinpoint exactly what went wrong.
About the author Sahil Khanna
Sahil Khanna is a Sr. Product Marketing Manager at Moogsoft, where he focuses on the emergence of Algorithmic IT Operations. In his free time, Sahil enjoys banging on drums and participating in high-stakes bets.