DevOpsPadDownload

Read Pod logs, including a crashed container

To read logs, list Pods with get pods, open a Pod and choose the Logs tab. Pick a container, choose how many lines to tail and, if the container restarted, turn on Previous instance to see the run that crashed. Each read is a snapshot with timestamps, not a live stream.

DevOpsPad does not store log output and never sends it to the AI assistant.

The Logs tab of the resource inspector with the container and tail-lines selectors, Previous instance checked, and a timestamped log line in the output.

Read a snapshot

  1. Run get pods (add a namespace if needed) and double-click the Pod.
  2. Choose Logs. You can also click a container name on the Overview tab to jump straight to its logs.
  3. In Container, pick the container. Each entry shows its type: Application, Init or Ephemeral.
  4. In Tail lines, choose 100, 200, 500, 1000 or 2000. The default is 200.
  5. Read the output. Copy logs puts the snapshot on the clipboard. Refresh logs fetches it again.

See the previous instance after a crash

When a container is in CrashLoopBackOff, its current instance often has no output yet. Check Previous instance to request the logs of the run before the last restart, the same as kubectl logs --previous. The Restarts column on the Overview tab tells you whether a previous instance exists.

If there is no previous instance, the cluster returns an error and DevOpsPad shows it in the tab.

Init and ephemeral containers

Init containers that failed are a common reason a Pod never starts. They appear in the Container list with the type Init. Ephemeral debug containers appear as Ephemeral.

Limits

  • At most 2,000 lines and 256 KiB per snapshot. When the size limit is hit, DevOpsPad says “Log size limit reached” and suggests fewer tail lines.
  • Requests time out after 20 seconds.
  • No live follow, no search inside the output and no download. Copy the text if you need to keep it.
  • “No log output returned for this container.” means the container wrote nothing in the requested range.

Permissions

Logs need get on pods and on the pods/log subresource in the Pod’s namespace. See Kubernetes permissions DevOpsPad needs. Permission errors stay inside the Logs tab.

Frequently asked questions

Can DevOpsPad follow logs live like kubectl logs -f?

Not yet. Each read is a snapshot of up to 2,000 lines and 256 KiB. Click Refresh logs to fetch the latest lines.

How do I see why a container crashed?

Select the container and turn on Previous instance. That shows the logs of the container run before the latest restart, which is usually where the crash message is.