Node level metrics question about prometheus-adapter HOT 8 CLOSED

Same issue here, still a problem? Seems I can't access any resources. Always Error from server (InternalError): Internal error occurred: unable to list matching resources

from prometheus-adapter.

i have some namespace prefixed metrics and cannot access them, same error.

{ "name": "namespaces/node_memory_PageTables_bytes", "singularName": "", "namespaced": false, "kind": "MetricValueList", "verbs": [ "get" ] }

from prometheus-adapter.

I am facing the same issue. Did you make any progress on this?

from prometheus-adapter.

can you guys please share the adapter config you are using to expose node metrics? I am getting the following error: server could not find the metric node_memory_Active_file for nodes

from prometheus-adapter.

Can you please post logs from the adapter (at high verbosity) and see if you find any related messages? Thanks!

from prometheus-adapter.

Namespace metic is special, we should get it like :

kubectl get --raw /apis/custom.metrics.k8s.io/v1beta1/namespaces/*/metrics/kube_pod_info

As for node metric, according to my test, it's ok to get it with the latest adapter code.

However there are very few node metrics with default config, because metrics from node exporter will not be labeled with node=XXX and the metrics' name usually begin with node_

If we want to get more node metrics, may be we could try to get metrics begin with node_ and override the instance to node.

@DirectXMan12 WDYT?

from prometheus-adapter.

I guess this is covered by #193, now present in the FAQ section.

from prometheus-adapter.

有有效的解决办法了吗

from prometheus-adapter.