Bring up kubernetes dashboard on master

[iampr...@gmail.com]

We are trying to setup a multinode kubernetes cluster with 1 master and multiple worker nodes in ubuntu 14.04.  --register-node=true and --register-schedulable=false is set while installing the kubernetes master so as to make the master not schedulable. Hence the kubernetes dashboard pod got scheduled on one of the available worker nodes. However we want the dashboard to be running on master.
 
As a work around started kubernetes dashboard as a docker container on master by using following command.
 
docker run --net=host --rm -it gcr.io/google_containers/kubernetes-dashboard-amd64:v1.4.0-beta1 --apiserver-host http://127.0.0.1:8080
 
Though the kubernetes dashboard is accessible, the page displayed internalerror and got refreshed automatically with the following error logs.

Incoming HTTP/1.1 GET /api/v1/workload/default?itemsPerPage=10&page=1 request from XXX.XX.XX.X
Getting lists of all workloads
Getting pod metrics
the server could not find the requested resource (get daemonsets.extensions)
[2016-08-31T03:08:56Z] Outcoming response to XXX.XX.XX.X with 500 status code

 
When I tried with a different dashboard image(livew.io/kubernetes-dashboard), I could successfully access kubernetes dashboard gui.
 
Can someone clarify the following.
 
1) Why the google kubernetes dashboard image didnt work in standalone docker container? Is their a different dashboard image that I need to pick?
2) Is their a better way to start kubernetes-dashboard GUI on master.


kubectl version
Client Version: version.Info{Major:"0", Minor:"19", GitVersion:"v0.19.3", GitCommit:"3103c8ca0f24514bc39b6e2b7d909bbf46af8d11", GitTreeState:"clean"}
Server Version: version.Info{Major:"0", Minor:"19", GitVersion:"v0.19.3", GitCommit:"3103c8ca0f24514bc39b6e2b7d909bbf46af8d11", GitTreeState:"clean"}

docker version

Client version: 1.6.2
Client API version: 1.18
Go version (client): go1.2.1
Git commit (client): 7c8fca2
OS/Arch (client): linux/amd64
Server version: 1.6.2
Server API version: 1.18
Go version (server): go1.2.1
Git commit (server): 7c8fca2
OS/Arch (server): linux/amd64

[Robert Bailey]

I'd be surprised if the dashboard worked on a 0.19.3 cluster. The first commit for the kubernetes dashboard was in October 2015 after Kubernetes 1.1 shipped. It was initially released alongside Kubernetes 1.2.0 (see the release notes). The error that you posted (not being able to find the daemonset type) is because that type didn't yet exist in 0.19.3 and the dashboard wasn't designed to be totally backwards compatible with all prior Kubernetes releases. 

On Wed, Aug 31, 2016 at 10:17 AM, <iampr...@gmail.com> wrote:

We are trying to setup a multinode kubernetes cluster with 1 master and multiple worker nodes in ubuntu 14.04.  --register-node=true and --register-schedulable=false is set while installing the kubernetes master so as to make the master not schedulable. Hence the kubernetes dashboard pod got scheduled on one of the available worker nodes. However we want the dashboard to be running on master.
 
As a work around started kubernetes dashboard as a docker container on master by using following command.
 
docker run --net=host --rm -it gcr.io/google_containers/kubernetes-dashboard-amd64:v1.4.0-beta1 --apiserver-host http://127.0.0.1:8080

Another option is to run the dashboard as a static pod by writing a manifest into the /etc/kubernetes/manifests directory (which the kubelet uses as another source of pods to run). 

 

 
Though the kubernetes dashboard is accessible, the page displayed internalerror and got refreshed automatically with the following error logs.

Incoming HTTP/1.1 GET /api/v1/workload/default?itemsPerPage=10&page=1 request from XXX.XX.XX.X
Getting lists of all workloads
Getting pod metrics
the server could not find the requested resource (get daemonsets.extensions)
[2016-08-31T03:08:56Z] Outcoming response to XXX.XX.XX.X with 500 status code

 
When I tried with a different dashboard image(livew.io/kubernetes-dashboard), I could successfully access kubernetes dashboard gui.
 
Can someone clarify the following.
 
1) Why the google kubernetes dashboard image didnt work in standalone docker container? Is their a different dashboard image that I need to pick?
2) Is their a better way to start kubernetes-dashboard GUI on master.


kubectl version
Client Version: version.Info{Major:"0", Minor:"19", GitVersion:"v0.19.3", GitCommit:"3103c8ca0f24514bc39b6e2b7d909bbf46af8d11", GitTreeState:"clean"}
Server Version: version.Info{Major:"0", Minor:"19", GitVersion:"v0.19.3", GitCommit:"3103c8ca0f24514bc39b6e2b7d909bbf46af8d11", GitTreeState:"clean"}

docker version

Client version: 1.6.2
Client API version: 1.18
Go version (client): go1.2.1
Git commit (client): 7c8fca2
OS/Arch (client): linux/amd64
Server version: 1.6.2
Server API version: 1.18
Go version (server): go1.2.1
Git commit (server): 7c8fca2
OS/Arch (server): linux/amd64

--
You received this message because you are subscribed to the Google Groups "Kubernetes developer/contributor discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kubernetes-dev+unsubscribe@googlegroups.com.
To post to this group, send email to kubernetes-dev@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/fbd25c94-c843-4333-bc25-78528fe78784%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Piotr Bryk]

On Tue, Sep 20, 2016 at 9:02 AM, Robert Bailey <robert...@google.com> wrote:

I'd be surprised if the dashboard worked on a 0.19.3 cluster. The first commit for the kubernetes dashboard was in October 2015 after Kubernetes 1.1 shipped. It was initially released alongside Kubernetes 1.2.0 (see the release notes). The error that you posted (not being able to find the daemonset type) is because that type didn't yet exist in 0.19.3 and the dashboard wasn't designed to be totally backwards compatible with all prior Kubernetes releases. 

Dashboard works with v1.2 and newer clusters. It should be easy to make it work with v0.19, but I'd strongly advise to update the cluster instead. You can contribute on file an issue on https://github.com/kubernetes/dashboard. 

--

Piotr Bryk 

 

To post to this group, send email to kuberne...@googlegroups.com.