Are you fascinated by Container technology? I am personally very impressed with it. How wonderful is it that now I don’t just ship the software, but I also ship its run-time environment in a self-contained autonomous system called Container? I can now get a good night’s sleep without worrying about unknown demons in the run-time environment bringing my software down. Goodbye nightmares of an angry customer eagerly waiting for me to tell how good a software developer I am.
The world seems to be moving away from all kinds of fixedness – functional and structural. Computing power and storage is abstracted in a cloud. Software defined networking devices morph in to a router or a switch at a moment’s notice. Virtualization has taken the world by storm, and as we were thinking what could be the next level of abstraction and virtualization – the beautiful idea of Containers came along. Docker is certainly the leader of that camp.
The flexibility and agility of IT resources make our life simpler and better. However, there is one person whose life got a little more difficult—the one who is responsible to monitor and ensure that this highly flexible, agile, fast and dynamically changing IT infrastructure is healthy. Those days of monitoring a well-known IT infrastructure that was set up with a well-known usage pattern are long gone. We are now in the age of unpredictability and virtualization; the Containers just added a newer level of abstraction and dynamicity to the world—bringing in a newer set of monitoring challenges.
Let me now come to the main topic of this blog, which I have titled “Dock your Docker”. This title conveys that Docker has arrived at your port and you can Dock it safely. You now have a monitoring solution that will take care of the automated monitoring of your Docker environment at your disposal.
The automated monitoring solution I am referring to is the OMi Management Pack for Docker (OMi MP for Docker). This solution automatically discovers the Docker Container instances running on a Linux node and monitors key facets in a single pane of glass.
These facets include:
Availability of Containers
Resource utilization metrics of each Container like CPU and Memory utilization, Page in and Page out details, Cache utilization etc.
Configurable threshold values for alerting on the monitored metrics
Support for data logging and graphing features
Docker implements APIs that enable the creation of containers that allow for running processes in isolation. Docker does not require an Operating System. It leverages the features of the Linux Operating System that allows for resource isolation and creation of namespaces. The processes running in a container get their own private view of the Operating System. While this approach provides easy creation of autonomous systems and enables rapid deployment automation, it is highly essential to monitor the Operating System level resources utilized by the Docker Containers themselves. The OMi Management Pack for Docker completely addresses this concern by providing automated monitoring of all the important metrics in this context.
As the saying goes, “…a picture is worth a thousand words…” - The picture below depicts a high-level Docker architecture and the Metrics monitored and Topology discovered by OMi MP for Docker.
The topology-driven monitoring enables you to easily visualize your managed environment and see the reported problems in the correct context. You can filter on the reported events by navigating the topology tree and selecting the instance of Node or Docker Container you are interested in. In a similar way you can also visualize performance graphs on any of the metrics logged for a specific monitored instance.
The OMi Management Pack for Docker is released and maintained under the Community model and can be downloaded for free here.
The OMi Management Pack for Docker is created using the OMi Management Pack Development Kit, which enables rapid authoring of monitoring solutions (OMi Management Packs) that can work on the HP OMi platform using simple Perl scripts and configuration files.
If you have any questions/views on this blog you write a comment below or you can post your questions and comments to the BSM Community Forum on the HP Live Network.