Among developers, Docker containers are clearly all the rage these days. But it may very well turn out to be the Docker API that winds up saving IT organizations from a management crisis that is starting to build inside IT operations teams.
As much as developers might love how simple it is to spin up a Docker container, it’s that very simplicity that creates a major IT operations challenge. Docker containers now spin up so quickly, IT organizations have little to no visibility into what’s actually occurring in the IT environment. To address that issue, Dynatrace this week announced a Docker plug-in that enabled Dynatrace Application Performance Monitoring (APM) software to collect data via the Docker API.
Martin Etmajer, a senior technology strategist for Dynatrace, told ProgrammableWeb that in addition to being able to see how much CPU and memory resources Docker applications are consuming, Dynatrace also provides the ability to monitor disk usage. As more Docker containers wind up being distributed throughout the enterprise, Etmajer notes that the potential for existing DevOps challenges to be further exacerbated becomes more pronounced.
The primary reason for this is that Docker containers are likely to be deployed on a mix of bare-metal servers, virtual machine and platform-as-a-service environments. As those Docker containers spin up and down, IT infrastructure contention issues become more challenging to manage. The Docker API provides a mechanism for unifying data collection across all these environments in a way that an APM tool can turn into actionable intelligence.
Etmajer says that Dynatrace works by injecting an agent into the Docker container environment that communicates with the Docker API. That data is then fed into the Dynatrace APM service to give IT organizations insight into the behavior of both the container and the host it is running on.
Thanks to broad industry support that is manifesting itself primarily in the form of a Cloud Native Computing Foundation led by The Linux Foundation, interoperability driven by the Docker API is advancing rapidly across multiple types of containers. The good news for developers is that rather than waiting for IT operations teams to allocate resources, developers can either spin Docker containers up on their own or access them via any number of APIs exposed by, for example, cloud service providers.
The challenge that developers would do well to remember is that on the other side of all those Docker containers is an IT operations teams that is struggling to keep up.