Merge pull request #234 from carmanzhang/monitor

Bug fix: pod count metrics

pkg/models/metrics/metricsruleconst.go

@@ -330,7 +330,7 @@ var RulePromQLTmplMap = MetricMap{
 	"cluster_namespace_count": `count(kube_namespace_annotations)`,
 
 	// cluster_pod_count = cluster_pod_running_count + cluster_pod_succeeded_count + cluster_pod_abnormal_count
-	"cluster_pod_count":           `sum(kube_pod_info unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0))`,
+	"cluster_pod_count":           `sum((kube_pod_status_scheduled{condition="true"} > 0)  * on (pod) group_left(node) kube_pod_info unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0))`,
 	"cluster_pod_quota":           `sum(kube_node_status_capacity_pods unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0))`,
 	"cluster_pod_utilisation":     `sum(kube_pod_info unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0)) / sum(kube_node_status_capacity_pods unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0))`,
 	"cluster_pod_running_count":   `count(kube_pod_info unless on (pod) (kube_pod_status_phase{phase=~"Failed|Pending|Unknown|Succeeded"} > 0) unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0))`,
@@ -405,7 +405,7 @@ var RulePromQLTmplMap = MetricMap{
 	"node_disk_inode_usage":       `node:disk_inodes_total:$1 - node:disk_inodes_free:$1`,
 	"node_disk_inode_utilisation": `(1 - (node:disk_inodes_free:$1 / node:disk_inodes_total:$1))`,
 
-	"node_pod_count":           `sum(kube_pod_info$1) by (node) unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0)`,
+	"node_pod_count":           `sum by (node) ((kube_pod_status_scheduled{condition="true"} > 0)  * on (pod) group_left(node) kube_pod_info$1 unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0))`,
 	"node_pod_quota":           `sum(kube_node_status_capacity_pods$1) by (node) unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0)`,
 	"node_pod_utilisation":     `(sum(kube_pod_info$1) by (node) / sum(kube_node_status_capacity_pods$1) by (node)) unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0)`,
 	"node_pod_running_count":   `count(kube_pod_info$1 unless on (pod) (kube_pod_status_phase{phase=~"Failed|Pending|Unknown|Succeeded"} > 0))  by (node) unless on (node) (kube_node_status_condition{condition="Ready",status=~"unknown|false"} > 0)`,
@@ -425,7 +425,7 @@ var RulePromQLTmplMap = MetricMap{
 	"namespace_memory_usage_wo_cache": `namespace:container_memory_usage_bytes_wo_cache:sum{namespace!="", namespace=~"$1"}* on (namespace) group_left(label_kubesphere_io_workspace)(kube_namespace_labels)`,
 	"namespace_net_bytes_transmitted": `sum by (namespace) (irate(container_network_transmit_bytes_total{namespace!="", namespace=~"$1", pod_name!="", ` + ExcludedVirtualNetworkInterfaces + `, job="kubelet"}[5m]))* on (namespace) group_left(label_kubesphere_io_workspace)(kube_namespace_labels)`,
 	"namespace_net_bytes_received":    `sum by (namespace) (irate(container_network_receive_bytes_total{namespace!="", namespace=~"$1", pod_name!="", ` + ExcludedVirtualNetworkInterfaces + `, job="kubelet"}[5m])) * on (namespace) group_left(label_kubesphere_io_workspace)(kube_namespace_labels)`,
-	"namespace_pod_count":             `sum(kube_pod_status_phase{namespace!="", namespace=~"$1"}) by (namespace) * on (namespace) group_left(label_kubesphere_io_workspace)(kube_namespace_labels)`,
+	"namespace_pod_count":             `sum(kube_pod_status_phase{phase!~"Failed|Succeeded", namespace!="", namespace=~"$1"}) by (namespace) * on (namespace) group_left(label_kubesphere_io_workspace)(kube_namespace_labels)`,
 	"namespace_pod_running_count":     `sum(kube_pod_status_phase{phase="Running", namespace!="", namespace=~"$1"}) by (namespace) * on (namespace) group_left(label_kubesphere_io_workspace)(kube_namespace_labels)`,
 	"namespace_pod_succeeded_count":   `sum(kube_pod_status_phase{phase="Succeeded", namespace!="", namespace=~"$1"}) by (namespace) * on (namespace) group_left(label_kubesphere_io_workspace)(kube_namespace_labels)`,
 	"namespace_pod_abnormal_count":    `sum(kube_pod_status_phase{phase=~"Failed|Pending|Unknown", namespace!="", namespace=~"$1"}) by (namespace) * on (namespace) group_left(label_kubesphere_io_workspace)(kube_namespace_labels)`,
@@ -536,7 +536,7 @@ var RulePromQLTmplMap = MetricMap{
 	"workspace_memory_usage_wo_cache": `sum(namespace:container_memory_usage_bytes_wo_cache:sum{namespace!="", namespace$1})`,
 	"workspace_net_bytes_transmitted": `sum(sum by (namespace) (irate(container_network_transmit_bytes_total{namespace!="", namespace$1, pod_name!="", ` + ExcludedVirtualNetworkInterfaces + `, job="kubelet"}[5m])))`,
 	"workspace_net_bytes_received":    `sum(sum by (namespace) (irate(container_network_receive_bytes_total{namespace!="", namespace$1, pod_name!="", ` + ExcludedVirtualNetworkInterfaces + `, job="kubelet"}[5m])))`,
-	"workspace_pod_count":             `sum(kube_pod_status_phase{namespace!="", namespace$1})`,
+	"workspace_pod_count":             `sum(kube_pod_status_phase{phase!~"Failed|Succeeded", namespace!="", namespace$1})`,
 	"workspace_pod_running_count":     `sum(kube_pod_status_phase{phase="Running", namespace!="", namespace$1})`,
 	"workspace_pod_succeeded_count":   `sum(kube_pod_status_phase{phase="Succeeded", namespace!="", namespace$1})`,
 	"workspace_pod_abnormal_count":    `sum(kube_pod_status_phase{phase=~"Failed|Pending|Unknown", namespace!="", namespace$1})`,