kubectl logs¶
Shows the output of the containers in a pod.

Syntax¶
kubectl logs [OPTIONS] POD | TYPE/NAME [-c CONTAINER]
Options¶
| Option | Description |
|---|---|
POD, deploy/NAME |
A pod, or a pod of this deployment |
-l LABEL=VALUE |
All pods with this label |
-f, --follow |
Keep showing new lines |
--tail N |
Only the last N lines |
--since 1h |
Only lines of the last hour (or 10m, 30s) |
-p, --previous |
Output of the previous, crashed container |
-c CONTAINER |
Container in a pod with several containers |
--prefix |
Put the pod and container name in front of every line |
Examples¶
The last lines of a deployment:
kubectl logs -n status deploy/web --tail 3
Found 2 pods, using pod/web-9bc5f7f5b-8ptw2
198.51.100.23 - - [26/Oct/2026:09:42:10 +0000] "GET / HTTP/1.1" 200 615 "-" "curl/8.14.1" "198.51.100.23"
198.51.100.23 - - [26/Oct/2026:09:42:12 +0000] "GET /status HTTP/1.1" 404 153 "-" "curl/8.14.1" "198.51.100.23"
198.51.100.23 - - [26/Oct/2026:09:42:15 +0000] "GET / HTTP/1.1" 200 615 "-" "curl/8.14.1" "198.51.100.23"
All pods of an application at once:
kubectl logs -n status -l app=web --prefix --tail 1
[pod/web-9bc5f7f5b-8ptw2/nginx] 198.51.100.23 - - [26/Oct/2026:09:42:15 +0000] "GET / HTTP/1.1" 200 615 "-" "curl/8.14.1" "198.51.100.23"
[pod/web-9bc5f7f5b-m8l96/nginx] 198.51.100.23 - - [26/Oct/2026:09:42:13 +0000] "GET /favicon.ico HTTP/1.1" 404 153 "-" "curl/8.14.1" "198.51.100.23"
Find out why the container crashed:
kubectl logs -n status web-7d4c9b8f6-xk2lp --previous
2026/10/26 09:40:02 [emerg] 1#1: unknown directive "sever_name" in /etc/nginx/conf.d/default.conf:3
nginx: [emerg] unknown directive "sever_name" in /etc/nginx/conf.d/default.conf:3
Follow live:
kubectl logs -n status -f deploy/web --tail 0
Found 2 pods, using pod/web-9bc5f7f5b-8ptw2
198.51.100.23 - - [26/Oct/2026:09:42:31 +0000] "GET / HTTP/1.1" 200 615 "-" "curl/8.14.1" "198.51.100.23"
^C
When do I need it?¶
- When a pod is in
CrashLoopBackOffand you need to know why - To follow the requests of an application while testing
- To check the output of a job or a migration
Logs of a crashed container
After a crash, kubectl logs shows the output of the new container, which is often
still empty. The reason for the crash is in kubectl logs --previous POD.
Logs disappear with the pod
When a pod is deleted or replaced, its logs are gone too. For anything you need to keep, collect the logs centrally, for example with Loki or the ELK stack.
Related commands¶
kubectl get · kubectl describe · kubectl events · kubectl exec