Want to have a single command to have continue streaming of logs from particular pod. Currently I have to do
kubectl get pods (lists all the running pods)kubectl logs <pod name> -f (continuous streaming of logs from pod)Bonus points: list pods starting with certain word like, kubectl get pods asset*. Which would just display pods with names starting with asset
You can use awk to achieve this. You can use:
kubectl logs -n <namespace> $(kubectl get pod -n <namespace> | awk '/<pattern>/{print $1}') -fHere, you can specify regex in <pattern> field. print $1 prints only first column of matched object. In our case, it is pod name.
For example:
kubectl logs -n kube-system $(kubectl get pod -n kube-system | awk '/kube-proxy*/{print $1}') -fYou can use some bash tricks to get that :
Example:
Get the logs for pod starting with name pattern core in kube-system namespace.
kubectl logs --namespace=kube-system $(kubectl get pods --namespace=kube-system -o go-template --template '{{range .items}}{{.metadata.name}}{{"\n"}}{{end}}' | grep core)or even simple:
kubectl logs --namespace=kube-system $(kubectl get pods -o=name --all-namespaces | grep core)but in production you should use fluentbit to stream the logs to ELK and then use filters in kibana UI to get the logs for each pod/deployment/namespace etc.
you can get all result by kubectl get pods and grep can do the rest of things.
[root@k8s-m1 ~]# kubectl get pods | grep httpd*-app
httpd-app-66cb7d499b-4nfl4 1/1 Running 0 6m39s
httpd-app-66cb7d499b-c2blq 1/1 Running 0 6m39s
httpd-app-66cb7d499b-dt4tr 1/1 Running 0 6m39sFinally was able to figure out the solution. This would be somewhat hacky, but I would basically use --field-selector=status.phase=Running and just get the name with -o=name flag. My final command would be something like
kubectl logs -f $(kubectl get pods --field-selector=status.phase=Running -o=name | awk '/asset/ {print $1;exit}')Links: Field Selectors