Error: could not find a ready tiller pod

[vielmetti]

I'm bringing up Kubernetes on my Mac under minikube and xhyve, with the eye towards using Kubernetic to manage the whole thing. I installed helm from homebrew as brew install kubernetes-helm.

When I do pretty much any Helm command, I get this error message:

$ helm status
Error: could not find a ready tiller pod

which seems pretty fundamental of a problem. I am not sure how to proceed from here, suggestions for debugging or component installation welcomed. However the upgrade seems to work:

$ helm init --upgrade
$HELM_HOME has been configured at /Users/emv/.helm.

Tiller (the helm server side component) has been upgraded to the current version.
Happy Helming!

[bacongobbler]

if you run kubectl -n kube-system get po, do you see a tiller-deploy pod available there?

[vielmetti]

I see two pods in mid-deploy:

Edwards-MacBook-Air:~ emv$ kubectl -n kube-system get po
NAME                             READY     STATUS              RESTARTS   AGE
kube-addon-manager-minikube      0/1       ContainerCreating   0          46m
tiller-deploy-2654728925-j2zvk   0/1       ContainerCreating   0          28m

Searching through kubectl logs I see this error:

E0303 22:15:59.462029    2423 docker_manager.go:2114] Failed to create pod infra container: ErrImagePull; Skipping pod "kube-addon-manager-minikube_kube-system(46ae05e07c52d84167b077b142aa4a39)": image pull failed for gcr.io/google_containers/pause-amd64:3.0, this may be because there are no credentials on this request.  details: (Error response from daemon: Get https://gcr.io/v1/_ping: dial tcp 173.194.66.82:443: i/o timeout)

which suggests that I missed something fundamental in setup....

[vielmetti]

I did two things, and now it's working.

minikube addons enable registry-creds seemed to address the problem of "no credentials on this request".

I also waited a long time, and it came up. I can't be certain that the registry-creds addon was the fix, but now it's working:

$ kubectl -n kube-system get po
NAME                             READY     STATUS    RESTARTS   AGE
kube-addon-manager-minikube      1/1       Running   0          1h
kube-dns-v20-ir9j5               3/3       Running   0          51m
kubernetes-dashboard-gxj81       1/1       Running   0          51m
tiller-deploy-2654728925-j2zvk   1/1       Running   0          1h

Thanks @bacongobbler for the debug help, I'll close this out.

[peterzandbergen]

@vielmetti I would like to know what command you used to find the mentioned log file entry. Cheers

[vielmetti]

@peterzandbergen I wish I knew which command I used to check the logs too - should have written that down when I did it.

[peterzandbergen]

Sounds familiar. Happens to me all the time. Thanks for replying.

[wethinkagile]

In my case Tiller was stuck booting. So I waited 10 minutes and then it worked. Would it be safe to delete the Tiller Pod if necessary? Or isnt there a ReplicaSet running?

[CharlieKuharski]

@vielmetti
Thanks! - I can confirm "minikube addons enable registry-creds" was the fix for my tiller pod not starting.

[charlenezheng]

Run this command may solve your issue.
kubectl logs --namespace kube-system tiller-deploy-2654728925-j2zvk

[kavemun]

hi, sorry to reopen this, in my case

$ kubectl get po -n kube-system
NAME                                  READY     STATUS     RESTARTS   AGE
tiller-deploy-6479f5f854-p7nsn   1/1       Running   0          1h

but when i run helm version i still get Error: could not find a ready tiller pod

$ helm version
Client: &version.Version{SemVer:"v2.11.0", GitCommit:"2e55dbe1fdb5fdb96b75ff144a339489417b146b", GitTreeState:"clean"}
Error: could not find a ready tiller pod

I am still running on kubernetes v1.8.13 though if it means something.

is there something i am still missing?
thanks

[bacongobbler]

I'd try using 2.11 with a newer version of Kubernetes. Kubernetes supports the current version + 1 release back, which at the time of 2.11 would be 1.10 and 1.11. 1.8 might be a little old. Give that a shot and see what happens.

[vineethvijay]

None was working for my case(there wasn't the pod scheduling).
This worked:

I ran tiller locally,

$ <bin>/tiller
[main] 2018/12/19 11:10:41 Starting Tiller v2.12.0 (tls=false)
[main] 2018/12/19 11:10:41 GRPC listening on :44134
[main] 2018/12/19 11:10:41 Probes listening on :44135
[main] 2018/12/19 11:10:41 Storage driver is ConfigMap
[main] 2018/12/19 11:10:41 Max history per release is 0
...
...

$ export HELM_HOST=localhost:44134
$ helm version # Should connect to localhost.

[sdejan12cc]

Funny, for me it worked when I did
minikube addons disable registry-creds
and then
minikube addons enable registry-creds

Previously I had registry creds enabled.

[ravibhattinkw]

I have the same problem
[a-iamdevuser ~]$ kubectl -n kube-system get pod
NAME READY STATUS RESTARTS AGE
coredns-7d77776957-6s7h2 0/1 Pending 0 6h
coredns-7d77776957-hcwf6 0/1 Pending 0 6h
tiller-deploy-5c5455c544-6gczg 0/1 Pending 0 2h
[a-iamdevuser@ ~]$ kubectl logs tiller-deploy-5c5455c544-6gczg
Error from server (NotFound): pods "tiller-deploy-5c5455c544-6gczg" not found
[a-iamdevuser@ ~]$ helm version
Client: &version.Version{SemVer:"v2.12.1", GitCommit:"02a47c7249b1fc6d8fd3b94e6b4babf9d818144e", GitTreeState:"clean"}
Error: could not find a ready tiller pod

[bacongobbler]

looks like everything in kube-system is in a pending state. Are you sure your cluster is healthy?

[loganxavier]

Having the same error on a fresh machine, opened a question about here.

[DNXie]

@charlenezheng I was trying your solution and ran into this error:

Error from server (BadRequest): container "tiller" in pod "tiller-deploy-c48485567-9tlpw" is waiting to start: trying and failing to pull image

[ebrucucen]

For me, my installation was missing tiller service account. On istio installation folder, running this solved my issue.

kubectl apply -f install/kubernetes/helm/helm-service-account.yaml   
helm init --service-account tiller

[stevegore]

If anyone else arrives at this point because you're scripting up a helm init followed by a helm upgrade --install, the good news is you can now use the wait command to wait for the pod to be ready before you do the upgrade. In the below I'm using labels to select the right pod:

kubectl -n kube-system wait --for=condition=Ready pod -l name=tiller --timeout=300s

[source-ram]

allow master to run pods:
kubectl taint nodes --all node-role.kubernetes.io/master-

[micseydel]

Nothing in this thread helped me with the issue, although restarting Docker fixed it.

[BiteFoo]

allow master to run pods:
kubectl taint nodes --all node-role.kubernetes.io/master-

thanks,it works for me.