Troubleshooting

Debug

This document gives you hints for diagnostics and solving issues, using the (hidden) subcommand debug.

Note it is technical and assumes you have some knowledge of how Kubernetes operates.

Watching

While installing, you can watch the installation (opening another terminal) with the command:

ops debug watch

Check that no pods will go in error while deploying.

Configuration

You can inspect the configuration with the ops debug subcommand

• API host: ops debug apihost

• Static Configuration: ops debug config.

• Current Status: ops debug status

• Runtimes: ops debug runtimes

• Load Balancer: ops debug lb

• Images: ops debug images

Logs

You can inspect logs with ops debug log subcommand. Logs you can show:

• operator: ops debug log operator (continuously: ops debug log foperator)

• controller: ops debug log controller (continuously: ops debug log fcontroller)

• database: ops debug log couchdb (continuously: ops debug log fcouchdb)

• certificate manager: ops debug log certman (continuously: ops debug log fcertmap)

Kubernetes

You can detect which Kubernetes are you using with:

ops debug detect

You can then inspect Kubernetes objects with:

• namespaces: ops debug kube ns

• nodes: ops debug kube nodes

• pod: ops debug kube pod

• services: ops debug kube svc

• users: ops debug kube users

You can enter a pod by name (use kube pod to find the name) with:

ops debug kube exec P=<pod-name>

Kubeconfig

Usually, ops uses a hidden kubeconfig so does not override your Kubernetes configuration.

If you want to go more in-depth and you are knowledgeable of Kubernetes, you can export the kubeconfig with ops debug export F=<file>.

You can overwrite your kubeconfig (be aware there is no backup) with ops debug export F=-.