IBM Cloud Kubernetes Service Review
June 08, 2020

IBM Cloud Kubernetes Service Review

Lorenzo Patera | TrustRadius Reviewer
Score 9 out of 10
Vetted Review
Verified User

Overall Satisfaction with IBM Cloud Kubernetes Service

We use IBM Cloud Kubernetes Service for research purposes. It is used by the whole Cloud services research team. It allows us to scale up easily our applications and to set up autoscale mechanisms when a condition is verified. We use IBM Cloud Kubernetes Service because it is really cost-effective and permits a precise billing based on the used resources.
  • Cost effective
  • Horizontal Scaling
  • Microservice Approach
  • Containerization
  • Overkill for simple applications
  • Learning curve slow
  • Not all software are native compatible
  • You can scale up horizontally by adding low-cost nodes to your cluster.
  • Once configured, it requires really simple configurations for new worker nodes.
  • The switching phase is not trivial, and sometimes requires additional NFS storage for sharing volumes.
In our experience, we typically use IBM Cloud Kubernetes from CLI. It is really powerful and permits us to manage the entire cluster with simple APIs. CLI tools can be installed also on dumb terminals so that no powerful and GUI furnished nodes are required for managing the Cluster. Sometimes we use also the web GUI, really intuitive and complete, but less flexible than CLI.
We have deployed only custom applications on IBM Cloud Kubernetes Service.
I used also Kubernetes on-premise installation. It permits you to install a custom Kubernetes cluster in your physical machines. The advantages are clearly the complete access to the machines via ssh and the possibility of customization. In the cons, the costs are really high for maintaining an entire infrastructure by yourself. For our purposes, in the end, we chose IBM Cloud Kubernetes Service, totally managed by IBM.
  • IBM Cloud Container Registry
I use IBM Cloud Container Registry for storage and sharing of my custom container images.
The GUI and the CLI permit total control on your cluster. Moreover, IBM Cloud Kubernetes Service tries to overcome the lack of an ssh shell by providing a virtual ssh shell on every container.
Online training is really an important resource for using these tools. IBM's help center is rich in useful information and tips. Also, external guides and tutorials are available (e.g. on youtube), but I followed only IBM ones and I had no difficulties.
We will renew our licence for sure. Our research is boosted with IBM Cloud Kubernetes Service. We can develop applications really fast.
The Microservice approach offered by IBM Cloud Kubernetes Service is well suited for applications that can be easily split into several parts that work independently one from each other. It is not easy to split correctly the software in multiple parts, but with Kubernetes, you can manage in real-time the deployment of your application, so some services can be deployed on the same physical machine.
It is not appropriate for really monolithic applications that require a lot of data in memory that is shared with multiple threads (e.g. AI applications).