The Elastisys Tech Blog

Search

GKE success story: From manually managed VMs to Kubernetes

Tempus has been on a journey over the past few months, from manually managing VMs to worry free operations with Google Kubernetes Engine (GKE). At Elastisys we are glad to have played a part in this process and successfully helped ease the transition. What made Tempus want to switch? What benefits did they see? And should you also consider GKE for your workloads? Read this success story to find out!

About Tempus

Tempus is a company that has built a system for pre-school presence management used by more than 50 municipalities in Sweden. It consists of multiple products that work together for simplifying every day life for administrators and teachers, as well as parents and children. Tempus has grown to become a market leader in this space thanks to their close cooperation with their customers and users.

Tempus has now been running their production environment in GKE for a few weeks so we took the chance and asked them what it was like.

What was the Problem?

First of all, what made them want to switch? Tempus started out developing and running their applications on bare virtual machines in a cloud. It is a nice quick way of getting started and playing around, but after a while it tends to get messy.

Tempus explains that it was time consuming to maintain the virtual machines. There were always updates to install. They also run into some performance issues with their old cloud provider. The cause for this was overloaded hardware, due to other co-located clients.

Tempus had already containerized their applications, but they were still doing the orchestration manually. They explain that administering the Docker containers was just too complex and time consuming, so they needed to fix this. And Kubernetes was a perfect match.

Solution

Kubernetes is built for orchestrating containerized workloads. It adds powerful abstractions and self healing capabilities that allows you to focus on your applications instead of moving containers around and connecting them to each other. Google’s Kubernetes Engine is a managed, production ready service. It allows you to deploy your applications in Kubernetes without having to install and manage your own cluster.

Even if it is easy to get started with GKE, it still takes some effort and knowledge to set up the environment. You often need to set up some external services, such as a database. And of course it also makes sense to automate this setup in case of problems later.

Elastisys helped Tempus in the transition by automating the provisioning of new Kubernetes clusters. In addition, we helped with manifest generation for the various components in the cluster. One of the trickiest parts in this case was the complicated ingress structure that Tempus required. But we managed to solve it and help Tempus get started with GKE.

After the handover, we had enough knowledge to continue with fine tuning of the development environment.

Up and Running!

Now Tempus is happily running their production workloads on GKE. There they don’t have to manage the operating system, network setting and low level applications any more. As a result, they are now free to spend time on more valuable tasks.

Now we only have to maintain the containers used to run our products.

It may be all nice and good when you are finally running in your new environment, but what about getting there? Tempus assures us that the transition went well. They have not had any major issues and they praise how smooth and easy it is to make changes in the environment. Some tweaking, balancing and optimization was needed initially. But this seems to have been a one time effort for settling in GKE.

What about observability then? GKE of course has its own tool for this: Stackdriver. However, Tempus has chosen to go with Datadog instead, for both monitoring and logging. They say that it was easy to get started with as it has built in support for Kubernetes.

Conclusion

GKE is a compelling option if you are interested in running your own Kubernetes cluster. It is easy to get started with and takes care of all the mundane tasks associated with managing your own virtual machines. As a result, you get more time for your precious applications.

If you need help or advice for transitioning to Kubernetes, don’t hesitate to reach out to us at Elastisys. As certified Kubernetes service partners, we have the knowledge you need!

P.S. Follow us on LinkedIn for more content like this.

 

Share:
LinkedIn
Twitter
Reddit