Enhance your ISTIO service mesh with IBM CKS
February 26, 2020

Enhance your ISTIO service mesh with IBM CKS

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

Overall Satisfaction with IBM Cloud Kubernetes Service

IBM Cloud Kubernetes Service is part of our testing environment for multi-cloud projects. By using ISTIO mesh to span to multiple Kubernetes clusters across multiple cloud providers, we can test the HA capabilities of the deployments and offer high HA solutions to our customers.
Also, by integrating IBM CKS to Google Cloud using Google Anthos GKE Connect, we can manage all our deployments in one place. IBM CKS offers one of the most powerful Kubernetes services: stable, powerful, easy to use and HIGH availability.
  • Easy to use and configure
  • ISTIO and KNATIVE installation by just clicking one button.
  • Use of CONTAINERD (CRI-O)
  • Choose the OS of the nodes
  • No additional training needed: IBM CKS works out of the box.
IBM's CKS GUI is not quite intuitive. In my opinion, it lacks a lot of useful information like performance or letting the users tune ISTIO configuration.
Using third-party services is one of the greatest advantages of IBM's CKS. Other managed K8s do not offer that.
OpenShift is AMAZING and the integration with IBM cloud is unique. OpenShift is more powerful than IBM Kubernetes for production workloads.
Well suited:
  • DEV / QA / SIT deployments: You can use IBMCKS for DEV / QA or SIT environments. Since IBM CKS is a managed service, you can let the deveolping team to be in charge of the infraestructure.
  • Multicloud deployments: Working with ISTIO mesh works perfect on IBM CKS.
  • Network intensive deployments: IBM's network is quite powerfull.
  • Docker containers
Less appropriate:
  • Production Workloads: Is never a good idea to use K8s for production if you don't have a team that fully understands how to operate K8s. K8s is one of the hardest techonologies to use due it's complexity.
  • Workloads that requieres KERNEL tunning: You can't modify Kernel parameters on Managed K8s like CKS. If you are running a workload like REDIS that needs certain KERNEL parameters to be set, I highly advice you to use K8s Open Source.
  • High performance workloads (GPU): IBM CKS does not offers GPUs on their nodes.

Using IBM Cloud Kubernetes Service

3 - Testing new products in order to evaluate adding them to the product portfolio.
1 - K8s needs very specific knowledge.
  • Deploy K8s clusters in Mexico! AMAZING!

Evaluating IBM Cloud Kubernetes Service and Competitors

  • Product Features
  • Vendor Reputation
  • Existing Relationship with the Vendor
  • Positive Sales Experience with the Vendor
Be in touch with a IBM vendor. We just began to use the product like a "normal" user. Maybe the next time I will look for assistance from an IBM vendor that can help us to fully undestand the capabilities of the product.

IBM Cloud Kubernetes Service Implementation

It was not the worst part but it was not the best.
Change management was minimal
  • Incomplete documentation
  • Basic support

IBM Cloud Kubernetes Service Training

  • Online training
  • no training
IBM's documentation is the WORST. It is very difficult to follow and very poorly designed.
NO, never. K8s is one the most difficult technologies to learn. I mean it!

IBM Cloud Kubernetes Service Support

The support is quite basic.
ProsCons
Problems get solved
No escalation required
Quick Initial Response
Slow Resolution
Less knowledgeable
Not kept informed
Need to explain problems multiple times
Support doesn't seem to care
No - K8s needs certified experts. IBM does not offer or guarantee that each of the Customer Service Agents is well trained. If I pay for Premium support, I need a confirmation from IBM that the agent is AT LEAST certified in K8s (CKA) and is not just a contact center agent that was hired to answer the chat/email.
My last interaction was through chat. It was very robotic and repetitive. I'm almost sure that I was chatting with WATSON instead of a human.

Using IBM Cloud Kubernetes Service

The GUI is very simple (in the good way) and easy to use.
ProsCons
Like to use
Relatively simple
Easy to use
Well integrated
Consistent
Feel confident using
Requires technical support
Lots to learn
  • ISTIO easy install
  • KNATIVE easy install
  • Kernel tunning
  • Automatic autoscaling

IBM Cloud Kubernetes Service Reliability

IBM's CKS does not offers automatic autoscaling nor vertical scaling (automatic). Other services like Google Kubernetes Engine scales up and down very well.
IBM's cloud is almost infallible.
IBM's cloud has a site in my conuntry (MEXICO) so the network latency was almost 0.