changes to hypershift metrics (#493)

* fine tuning some metric * enabled prom container usage
cloud-bulldozer · Oct 12, 2022 · 0367931 · 0367931
1 parent 60422fd
commit 0367931
Showing 1 changed file with 7 additions and 7 deletions.
diff --git a/workloads/kube-burner/metrics-profiles/hypershift-metrics.yaml b/workloads/kube-burner/metrics-profiles/hypershift-metrics.yaml
@@ -48,10 +48,10 @@
   metricName: podMemoryReq
   instant: true
 
-- query: (sum(irate(container_cpu_usage_seconds_total{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",name!="",container!="POD",namespace=~"openshift-(etcd|.*apiserver|ovn-kubernetes|sdn|ingress|.*controller-manager|.*scheduler|image-registry)"}[2m]) * 100) by (container, pod, namespace, node, openshift_cluster_name)) > 0
+- query: (sum(irate(container_cpu_usage_seconds_total{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",name!="",container!="POD",namespace=~"openshift-(etcd|.*apiserver|ovn-kubernetes|sdn|ingress|.*controller-manager|.*scheduler|image-registry|monitoring|user-workload-monitoring)"}[2m]) * 100) by (container, pod, namespace, node, openshift_cluster_name)) > 0
   metricName: mgmt-containerCPU
 
-- query: sum(container_memory_rss{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",name!="",container!="POD",namespace=~"openshift-(etcd|.*apiserver|ovn-kubernetes|sdn|ingress|.*controller-manager|.*scheduler|image-registry)"}) by (container, pod, namespace, node, openshift_cluster_name)
+- query: sum(container_memory_rss{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",name!="",container!="POD",namespace=~"openshift-(etcd|.*apiserver|ovn-kubernetes|sdn|ingress|.*controller-manager|.*scheduler|image-registry|monitoring|user-workload-monitoring)"}) by (container, pod, namespace, node, openshift_cluster_name)
   metricName: mgmt-containerMemory
 
 # Containers & pod metrics
@@ -90,17 +90,17 @@
 
 # Node metrics: CPU & Memory
 
-- query: (sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"}[2m])) by (mode,instance) and on (instance) label_replace(bottomk(1, min_over_time(irate(node_cpu_seconds_total{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}",mode=~"idle"}[2m])[24h:1m] @ ${Q_TIME})), "instance", "$1", "instance", "(.+)")) > 0
+- query: (sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"}[2m])) by (mode,instance) and on (instance) label_replace(bottomk(1, min_over_time(sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}",mode=~"idle"}[2m])) by (mode,instance)[5m:2m] @ ${Q_TIME})), "instance", "$1", "instance", "(.+)")) > 0
   metricName: nodeCPU-Workers
 
 - query: node_memory_MemAvailable_bytes{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"} and on (instance) label_replace(kube_node_role{role="worker"}, "instance", "$1", "node", "(.+)")
   metricName: nodeMemoryAvailable-Workers
 
 # Management Node metrics: CPU & Memory
-- query: (max((sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}"}[2m])) by (mode,instance) and on (instance) label_replace(bottomk(1, min_over_time(irate(node_cpu_seconds_total{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",mode=~"idle",instance!~"${MGMT_WORKER_ONLY_NODES}"}[2m])[24h:1m] @ ${Q_TIME})), "instance", "$1", "instance", "(.+)"))) by (mode, instance)) > 0
+- query: (max((sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}"}[2m])) by (mode,instance) and on (instance) label_replace(bottomk(1, min_over_time(sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",mode=~"idle",instance!~"${MGMT_WORKER_ONLY_NODES}"}[2m])) by (mode,instance)[5m:2m] @ ${Q_TIME})), "instance", "$1", "instance", "(.+)"))) by (mode, instance)) > 0
   metricName: mgmtNodeCPU-AggregatedWorkers
 
-- query: bottomk(1,min_over_time(node_memory_MemAvailable_bytes{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",instance!~"${MGMT_WORKER_ONLY_NODES}"}[24h:1m] @ ${Q_TIME}))
+- query: bottomk(1,min_over_time(node_memory_MemAvailable_bytes{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",instance!~"${MGMT_WORKER_ONLY_NODES}"}[5m:2m] @ ${Q_TIME}))
   metricName: mgmtNodeMemoryAvailable-AggregatedWorkers
 
 - query: (avg(node_memory_MemTotal_bytes{openshift_cluster_name=~"${MGMT_CLUSTER_NAME}",instance!~"${MGMT_WORKER_ONLY_NODES}"}) by (instance))
@@ -125,10 +125,10 @@
 - query: node_memory_MemTotal_bytes{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"} and on (instance) label_replace(kube_node_role{role="worker"}, "instance", "$1", "node", "(.+)")
   metricName: nodeMemoryTotal-Workers
 
-- query: (max(sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"}[2m])) by (mode,instance) and on (instance) label_replace(bottomk(1, min_over_time(irate(node_cpu_seconds_total{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}",mode=~"idle"}[2m])[24h:1m] @ ${Q_TIME})), "instance", "$1", "instance", "(.+)")) by (mode)) > 0
+- query: (max(sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"}[2m])) by (mode,instance) and on (instance) label_replace(bottomk(1, min_over_time(sum(irate(node_cpu_seconds_total{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}",mode=~"idle"}[2m])) by (mode,instance)[5m:2m] @ ${Q_TIME})), "instance", "$1", "instance", "(.+)")) by (mode)) > 0
   metricName: nodeCPU-AggregatedWorkers
 
-- query: bottomk(1,min_over_time(node_memory_MemAvailable_bytes{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"}[24h:1m] @ ${Q_TIME}))
+- query: bottomk(1,min_over_time(node_memory_MemAvailable_bytes{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"}[5m:2m] @ ${Q_TIME}))
   metricName: nodeMemoryAvailable-AggregatedWorkers
 
 - query: avg(node_memory_MemTotal_bytes{openshift_cluster_name=~"${HOSTED_CLUSTER_NAME}"} and on (instance) label_replace(kube_node_role{role="worker"}, "instance", "$1", "node", "(.+)"))