confusing about the currently resource which descheduler is collecting

[wang-xiaowu]

i know desceduler currently only collect request resource of pods，not actual used resource now i did not config any resource for my business pods，so what's the meaning of usagePercentage which desceduler's log shows this is descheduler's log

2023-09-08T16:21:37.043751128+08:00 I0908 08:21:37.043614       1 node.go:49] "Node lister returned empty list, now fetch directly"
I0908 08:21:37.050181       1 descheduler.go:278] Building a pod evictor
2023-09-08T16:21:37.050711423+08:00 I0908 08:21:37.050636       1 nodeutilization.go:227] "Node is appropriately utilized" node="k3s-release-node1" usage=map[cpu:2225m memory:2029Mi pods:11] usagePercentage=map[cpu:27.8125 memory:13.08786860528167 pods:3.6666666666666665]
2023-09-08T16:21:37.050724668+08:00 I0908 08:21:37.050664       1 nodeutilization.go:227] "Node is appropriately utilized" node="k3s-release-node2" usage=map[cpu:3825m memory:7693Mi pods:57] usagePercentage=map[cpu:47.8125 memory:49.661356515599486 pods:19]
2023-09-08T16:21:37.050733672+08:00 I0908 08:21:37.050680       1 nodeutilization.go:227] "Node is appropriately utilized" node="k3s-release-server1" usage=map[cpu:1525m memory:6279Mi pods:31] usagePercentage=map[cpu:19.0625 memory:40.50203225248057 pods:10.333333333333334]
2023-09-08T16:21:37.050737541+08:00 I0908 08:21:37.050702       1 nodeutilization.go:227] "Node is appropriately utilized" node="k3s-release-server2" usage=map[cpu:5325m memory:7753Mi pods:54] usagePercentage=map[cpu:66.5625 memory:50.00988976128762 pods:18]
2023-09-08T16:21:37.050745950+08:00 I0908 08:21:37.050717       1 nodeutilization.go:227] "Node is appropriately utilized" node="k3s-release-server3" usage=map[cpu:4135m memory:8869Mi pods:39] usagePercentage=map[cpu:51.6875 memory:57.20852731753642 pods:13]
2023-09-08T16:21:37.050750034+08:00 I0908 08:21:37.050736       1 lownodeutilization.go:118] "Criteria for a node under utilization" CPU=20 Mem=20 Pods=20
2023-09-08T16:21:37.050804680+08:00 I0908 08:21:37.050746       1 lownodeutilization.go:119] "Number of underutilized nodes" totalNumber=0
I0908 08:21:37.050772       1 lownodeutilization.go:132] "Criteria for a node above target utilization" CPU=80 Mem=80 Pods=50
2023-09-08T16:21:37.050826135+08:00 I0908 08:21:37.050781       1 lownodeutilization.go:133] "Number of overutilized nodes" totalNumber=0
I0908 08:21:37.050800       1 lownodeutilization.go:136] "No node is underutilized, nothing to do here, you might tune your thresholds further"

[a7i]

Hi @wang-xiaowu would you be able to share the output of kubectl describe node k3s-release-node1?

[wang-xiaowu]

# kubectl describe node k3s-release-node1
Name:               k3s-release-node1
Roles:              worker
Labels:             beta.kubernetes.io/arch=amd64
                    beta.kubernetes.io/instance-type=k3s
                    beta.kubernetes.io/os=linux
                    kubernetes.io/arch=amd64
                    kubernetes.io/hostname=k3s-release-node1
                    kubernetes.io/os=linux
                    node-role.kubernetes.io/worker=worker
                    node.kubernetes.io/instance-type=k3s
Annotations:        flannel.alpha.coreos.com/backend-data: {"VNI":1,"VtepMAC":"4a:f9:65:cc:c3:51"}
                    flannel.alpha.coreos.com/backend-type: vxlan
                    flannel.alpha.coreos.com/kube-subnet-manager: true
                    flannel.alpha.coreos.com/public-ip: 10.1.4.15
                    k3s.io/hostname: k3s-release-node1
                    k3s.io/internal-ip: 10.1.4.15
                    k3s.io/node-args: ["agent","--kubelet-arg","max-pods=300","eviction-hard=memory.available\u003c500Mi","system-reserved=memory=1Gi"]
                    k3s.io/node-config-hash: GGMEOSIPX3TR6JIYAUZOARYGMFGIHQZ6JAQ46UKNF46HI2SWHI4Q====
                    k3s.io/node-env:
                      {"K3S_DATA_DIR":"/var/lib/rancher/k3s/data/484bf694b486d93cc93bcf90f74d5c77628550d4456f21760fd720b88c93881e","K3S_TOKEN":"********","K3S_U...
                    management.cattle.io/pod-limits: {"cpu":"4400m","memory":"10144Mi"}
                    management.cattle.io/pod-requests: {"cpu":"2225m","memory":"2029Mi","pods":"11"}
                    node.alpha.kubernetes.io/ttl: 0
                    volumes.kubernetes.io/controller-managed-attach-detach: true
CreationTimestamp:  Fri, 18 Nov 2022 14:02:38 +0800
Taints:             deploy=node1:NoSchedule
Unschedulable:      false
Lease:
  HolderIdentity:  k3s-release-node1
  AcquireTime:     <unset>
  RenewTime:       Wed, 13 Sep 2023 10:36:51 +0800
Conditions:
  Type             Status  LastHeartbeatTime                 LastTransitionTime                Reason                       Message
  ----             ------  -----------------                 ------------------                ------                       -------
  MemoryPressure   False   Wed, 13 Sep 2023 10:34:11 +0800   Wed, 17 May 2023 12:41:47 +0800   KubeletHasSufficientMemory   kubelet has sufficient memory available
  DiskPressure     False   Wed, 13 Sep 2023 10:34:11 +0800   Wed, 17 May 2023 12:41:47 +0800   KubeletHasNoDiskPressure     kubelet has no disk pressure
  PIDPressure      False   Wed, 13 Sep 2023 10:34:11 +0800   Wed, 17 May 2023 12:41:47 +0800   KubeletHasSufficientPID      kubelet has sufficient PID available
  Ready            True    Wed, 13 Sep 2023 10:34:11 +0800   Tue, 29 Aug 2023 10:57:32 +0800   KubeletReady                 kubelet is posting ready status
Addresses:
  InternalIP:  10.1.4.15
  Hostname:    k3s-release-node1
Capacity:
  cpu:                8
  ephemeral-storage:  103019184Ki
  hugepages-1Gi:      0
  hugepages-2Mi:      0
  memory:             15874976Ki
  pods:               300
Allocatable:
  cpu:                8
  ephemeral-storage:  100217062117
  hugepages-1Gi:      0
  hugepages-2Mi:      0
  memory:             15874976Ki
  pods:               300
System Info:
  Machine ID:                 1812ee08555f4a7fbedd6f763d53f359
  System UUID:                1812ee08-555f-4a7f-bedd-6f763d53f359
  Boot ID:                    5dd6f132-0dcb-4e25-a42c-4d54aea0a172
  Kernel Version:             5.4.224-1.el7.elrepo.x86_64
  OS Image:                   CentOS Linux 7 (Core)
  Operating System:           linux
  Architecture:               amd64
  Container Runtime Version:  containerd://1.7.1-k3s1
  Kubelet Version:            v1.24.16+k3s1
  Kube-Proxy Version:         v1.24.16+k3s1
PodCIDR:                      10.42.3.0/24
PodCIDRs:                     10.42.3.0/24
ProviderID:                   k3s://k3s-release-node1
Non-terminated Pods:          (11 in total)
  Namespace                   Name                                             CPU Requests  CPU Limits   Memory Requests  Memory Limits  Age
  ---------                   ----                                             ------------  ----------   ---------------  -------------  ---
  kube-system                 kube-state-metrics-64dcf75c88-lzgfz              0 (0%)        0 (0%)       0 (0%)           0 (0%)         118d
  kube-system                 node-local-dns-kzsgn                             25m (0%)      0 (0%)       5Mi (0%)         0 (0%)         298d
  sopei-biz                   wecom-robot-server-6b979cd497-8wc2r              100m (1%)     1 (12%)      100Mi (0%)       1Gi (6%)       13d
  sopei-biz                   wecom-robot-server-6b979cd497-fxmnx              100m (1%)     1 (12%)      100Mi (0%)       1Gi (6%)       13d
  sopei-biz                   xxl-job-admin-7f76749b7-hhcj8                    0 (0%)        0 (0%)       200Mi (1%)       1000Mi (6%)    118d
  sopei-biz                   xxl-job-executor-calculation-75946d5cf8-5lxmn    600m (7%)     800m (10%)   200Mi (1%)       1000Mi (6%)    32d
  sopei-biz                   xxl-job-executor-collection-6f5f5bb65c-hvzsq     500m (6%)     500m (6%)    200Mi (1%)       1000Mi (6%)    118d
  sopei-biz                   xxl-job-executor-sccs-69f7bf9d49-qd595           900m (11%)    1100m (13%)  200Mi (1%)       1000Mi (6%)    103d
  sopei-utils                 filebeat-ccdfx                                   0 (0%)        0 (0%)       0 (0%)           0 (0%)         19d
  sopei-utils                 kie-server-release-sccs-77797c6754-pz5x5         0 (0%)        0 (0%)       512Mi (3%)       2Gi (13%)      20d
  sopei-utils                 kie-server-release-sccs-77797c6754-z6rdp         0 (0%)        0 (0%)       512Mi (3%)       2Gi (13%)      20d
Allocated resources:
  (Total limits may be over 100 percent, i.e., overcommitted.)
  Resource           Requests      Limits
  --------           --------      ------
  cpu                2225m (27%)   4400m (55%)
  memory             2029Mi (13%)  10144Mi (65%)
  ephemeral-storage  0 (0%)        0 (0%)
  hugepages-1Gi      0 (0%)        0 (0%)
  hugepages-2Mi      0 (0%)        0 (0%)
Events:              <none>

[a7i]

Given the output of Descheduler:

"Node is appropriately utilized" node="k3s-release-node1" usage=map[cpu:2225m memory:2029Mi pods:11] usagePercentage=map[cpu:27.8125 memory:13.08786860528167 pods:3.6666666666666665]

And the node output showing that some pods (from namespace sopei-biz) have Requests defined, this seems to all match up, correct?

[a7i]

/kind support

[wang-xiaowu]

Given the output of Descheduler:

"Node is appropriately utilized" node="k3s-release-node1" usage=map[cpu:2225m memory:2029Mi pods:11] usagePercentage=map[cpu:27.8125 memory:13.08786860528167 pods:3.6666666666666665]

And the node output showing that some pods (from namespace sopei-biz) have Requests defined, this seems to all match up, correct?

yes so usage only shows the used resource of the pod which was defined requests，right？

[a7i]

Exactly, because that's what Scheduler uses

[wang-xiaowu]

i see thx~