What is node pool and how to set it about kube-eagle HOT 20 CLOSED

So you get everything, only node usage does not work.

Do you get reasonable output with this command?
kubectl top node

from kube-eagle.

Check logs of metrics-server. As soon as kubectl top node works, this project will work too I think.

from kube-eagle.

Here is what I have:

Nodes a loaded OK but graphs are empty. Still debugging and see why.

from kube-eagle.

btw, I'm running this in DigitalOcean k8s cluster, which also does have node pool thing.

from kube-eagle.

OK, tests some individual query in a chart and seems to work OK:

And I see that node pool is just some label from node name, no big deal here.

Prometheus is not my default datasource, can this be the issue ?

from kube-eagle.

Yes, that was it. Prometheus is not the default data source. Can we do smth about this? I guess each chart needs to include specific data source name.

from kube-eagle.

@botzill its not working for me , my logs are giving me the below error for the pod:
{"level":"info","msg":"Listening on 0.0.0.0:8080","time":"2019-03-04T13:36:18Z"}
{"level":"info","msg":"Creating InCluster config to communicate with Kubernetes master","time":"2019-03-04T13:36:18Z"}
panic: the server could not find the requested resource (get pods.metrics.k8s.io)

goroutine 7 [running]:
github.com/google-cloud-tools/kube-eagle/vendor/github.com/weeco/kube-eagle/pkg/metrics_store.Collect()
/go/src/github.com/google-cloud-tools/kube-eagle/vendor/github.com/weeco/kube-eagle/pkg/metrics_store/metrics_store.go:59 +0x386
main.main.func1()
/go/src/github.com/google-cloud-tools/kube-eagle/main.go:56 +0x32
created by main.main
/go/src/github.com/google-cloud-tools/kube-eagle/main.go:52 +0x47

any idea ??

from kube-eagle.

For me works OK this part, not sure what's wrong.

@weeco can tell us more.

from kube-eagle.

@saiyam1814 I think you need to install metrics-server. This is the new extension API that replaces Heapster.

For the datasource Issue, I would suggest a variable that selects all Prometheus Datasources and present them as a drop-down. Similar tho this:

from kube-eagle.

@saiyam1814 yes, that would be great.

from kube-eagle.

@botzill @towolf I have installe dmetrics server and my pod log says
{"level":"info","msg":"Creating InCluster config to communicate with Kubernetes master","time":"2019-03-04T14:23:34Z"}
{"level":"info","msg":"Listening on 0.0.0.0:8080","time":"2019-03-04T14:23:34Z"}
but when I visit http://127.0.0.1:8080/ its site cannot be reached and also there is no data in grafana dashboard

from kube-eagle.

@saiyam1814

Does this work for you?
kubectl get --raw /apis/metrics.k8s.io/v1beta1

If not, you need to set up metrics-server properly.

from kube-eagle.

@towolf yes
./kubectl.exe get --raw /apis/metrics.k8s.io/v1beta1
{"kind":"APIResourceList","apiVersion":"v1","groupVersion":"metrics.k8s.io/v1beta1","resources":[{"name":"nodes","singularName":"","namespaced":false,"kind":"NodeMetrics","verbs":["get","list"]},{"name":"pods","singularName":"","namespaced":true,"kind":"PodMetrics","verbs":["get","list"]}]}

and the dashboard do not show the graphs ...

but it shows the pods

from kube-eagle.

nope metrics not available for that @towolf

from kube-eagle.

@towolf hmmmm somehow its not giving any metrics ....
nable to fetch pod metrics for pod sample-weblogic-operator-pss/weblogic-operator-85c5b6c4d4-ldlh7: no metrics known for pod
E0304 14:55:09.168635 1 reststorage.go:144] unable to fetch pod metrics for pod default/upload-app-6b9b6467f7-jfxlw: no metrics known for pod
E0304 14:55:09.168643 1 reststorage.go:144] unable to fetch pod metrics for pod traefik/traefik-operator-75d97bc97b-gqwhf: no metrics known for pod
E0304 14:55:09.168651 1 reststorage.go:144] unable to fetch pod metrics for pod default/subserver-app-fbf679d4-zqskf: no metrics known for pod
E0304 14:55:09.177528 1 reststorage.go:129] unable to fetch n

from kube-eagle.

@saiyam1814 check my solution for that here: #9 (comment)

from kube-eagle.

@alfonmga yes worls fine now ... I am able see the metrics . But I dont know after sometime ... kube-eagle automatically stops .... need to check its behaviour over the day today. will update you Also If I import the dashboard from :
https://github.com/linuxacademy/content-kubernetes-prometheus-env/blob/master/grafana/dashboard/Kubernetes%20All%20Nodes.json
the results actually differ for cpu and memory usage
@towolf @weeco ^^

from kube-eagle.

@botzill I want to keep the grafana dashboard as generic as possibe. I believe therefore it's better to rely on prometheus being the default data source. Is there a better way to ensure a prometheus datasource in an exported grafana dashboard?

@saiyam1814 Your posted logs here indicate that you are not using the latest version of Kube Eagle. Maybe you haven't changed the default image tage provided in the helm chart. Please create a separate issue regarding the random "stops" (logs would help to figure that out).

from kube-eagle.

Well, we can have different setups and I think there should be a nice way to set Prometheus data source as a variable. On my setup Postgres was default and I had to change it and reload grafana. When things work out of the box, I like them :).

from kube-eagle.

Since the initial issue with node_pool variables is solved I'll go ahead and close this issue.

from kube-eagle.