IKS Diagnostics and Debug Tool

We found while responding to issues and helping our IBM Cloud Kubernetes Service (IKS) Customers that they appreciate the faster resolution to problems by giving them self-serve tools for figuring out what may have gone wrong. Also the gathered data helps our support to identify issues faster and cut back-and-forwards down.

We released a diagnostics and debug tool for IKS to speed up finding potential issues/problems in IBM Cloud’s managed Kubernetes service. You can read more details in my announcement.

I don’t want tiller in my cluster

Hearing this became very quickly a pattern.

You can use helm to deploy the IKS Diagnostics and Debug Tool in your cluster today. Although you need to have helm installed on your laptop, you don’t necessarily have to install tiller onto your cluster. Here is how.

Download the chart into ibm-diagnostics-tool-chart folder:

$ helm repo update
[...]
$ helm fetch ibm/ibmcloud-iks-debug --untar --untardir ./ibm-diagnostics-tool-chart

Assuming you have kubectl set up for your cluster, by running the following the helm will generate the .yamls that are needed, which you apply directly to your cluster. No tiller installed:

$ helm template --namespace ibm-system --name my-debugger ./ibm-diagnostics-tool-chart/ibmcloud-iks-debug/  | kubectl apply -f -

[It will create service account, role binding, daemonset, the debug tool itself.]

From here you can just follow the documented steps, like:

$ kubectl proxy --port 8080

and then open the debug tool UI in your browser:

$ open http://localhost:8080/api/v1/namespaces/ibm-system/services/my-debugger-ibm-iks-debug:8822/proxy/page

Voila, you have your tool running. :)

How can I delete it?

You run the same command, just with the delete command instead of apply

$ helm template --namespace ibm-system --name my-debugger ./ibm-diagnostics-tool-chart/ibm-iks-debug/  | kubectl delete -f -

This will remove everything you installed for the tool in your cluster.

Contact us

If you have questions, engage our team via Slack by registering here and join the discussion in the #general channel on our public IBM Cloud Kubernetes Service Slack.