search

kubernetes-sigs / scheduler-plugins

Repository for out-of-tree scheduler plugins based on scheduler framework.
Apache License 2.0
1.13k stars 523 forks source link

[trimaran] Cpu metric not found in node metrics #702

Open namgizlat opened 9 months ago

namgizlat commented 9 months ago

Area

Other components

No response

What happened?

while using the plugin trimaran following error appears in log: E0212 06:47:01.656601 1 targetloadpacking.go:142] "Cpu metric not found in node metrics" nodeName="node01.namgiz.com" nodeMetrics=[{Name:instance:node_memory_utilisation:ratio Type:Memory Operator:AVG Rollup:15m Value:11.70095792210995} {Name:instance:node_network_transmit_bytes:rate:sum Type:Bandwidth Operator:AVG Rollup:15m Value:8.235324565656565e+06} {Name:instance:node_network_receive_bytes:rate:sum Type:Bandwidth Operator:AVG Rollup:15m Value:1.6781439838383835e+07} {Name:instance:node_memory_utilisation:ratio Type:Memory Operator:STD Rollup:15m Value:0.03558027903707252} {Name:instance:node_network_transmit_bytes:rate:sum Type:Bandwidth Operator:STD Rollup:15m Value:502076.00473060855} {Name:instance:node_network_receive_bytes:rate:sum Type:Bandwidth Operator:STD Rollup:15m Value:3.155026755193178e+06}]

What did you expect to happen?

trimaran will use cpu metric for scheduling pods

How can we reproduce it (as minimally and precisely as possible)?

No response

Anything else we need to know?

Running on openshift (Version below)

Kubernetes version

``` Openshift Version: 4.12.40 Kubernetes Version: v1.25.14+20cda61 ```

Scheduler Plugins version

registry.k8s.io/scheduler-plugins/kube-scheduler:v0.27.8
k8s-triage-robot commented 6 months ago

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

You can:

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

wangchen615 commented 5 months ago

@Huang-Wei this issue is due to some metric name upgrades specific to OpenShift. We have made some changes in load watcher so it supports both OpenShift and Kubernetes, should we push an update on the versions of load watcher for Trimaran?

Huang-Wei commented 5 months ago

should we push an update on the versions of load watcher for Trimaran?

Yes, please. And come up with a release note detailing what metrics are both supported.

k8s-triage-robot commented 4 months ago

The Kubernetes project currently lacks enough active contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

You can:

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle rotten

namgizlat commented 3 months ago

@Huang-Wei this issue is due to some metric name upgrades specific to OpenShift. We have made some changes in load watcher so it supports both OpenShift and Kubernetes, should we push an update on the versions of load watcher for Trimaran?

Hi,

was it done?

Huang-Wei commented 3 months ago

cc @wangchen615

/remove-lifecycle rotten

k8s-triage-robot commented 2 weeks ago

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

You can:

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale