一、命名空间
Namespace = QCE/TKE2
二、监控指标
| 指标英文名 | 指标中文名 | 单位 | 维度 | 统计规则 [period, statType] |
| K8sNodeCpuUsage | CPU 利用率 | % | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeStatusReady | Node 状态 | – | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, min] [300s, min] [3600s, min] [86400s, min] |
| K8sNodeLanIntraffic | 内网入带宽 | MB/S | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeTcpCurrEstab | TCP 连接数 | 个 | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeLanOuttraffic | 内网出带宽 | MB/S | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeWanIntraffic | 外网入带宽 | MB/S | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeMemUsage | 内存利用率 | % | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodePodRestartTotal | 节点上 Pod 重启次数 | 次 | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, sum] [300s, sum] [3600s, sum] [86400s, sum] |
| K8sNodeCpuCoreRequestTotal | CPU 分配量 | 核 | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeGpuMemoryUsedBytes | GPU 内存使用量 | B | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeGpuUsed | GPU 利用率 | % | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeRateGpuMemoryUsed | GPU 内存利用率 | % | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeWanOuttraffic | 外网出带宽 | MB/S | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeRateGpuUsed | GPU 利用率 | % | 必填维度:tke_cluster_instance_id非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeMemoryRequestBytesTotal | 内存分配量 | MB | 必填维度:tke_cluster_instance_id、非必选:node_role、node、un_instance_id | [60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeAllocatedIpTotal | GlobalRouter 模式集群中节点 PodCIDR 已分配的 IP 数 | – | 必填维度:tke_cluster_instance_id 非必选:node_role、node、un_instance_id |
[60s, last] [300s, last] [3600s, last] [86400s, last] |
| K8sNodeDirectEniRequestTotal | node 的 direct-eni 分配量 | – | 必填维度:tke_cluster_instance_id 非必选:node_role、node、un_instance_id |
[60s, last] [300s, last] [3600s, last] [86400s, last] |
| K8sNodeEniIpRequestTotal | node 的 eni-ip 分配量 | – | 必填维度:tke_cluster_instance_id 非必选:node_role、node、un_instance_id |
[60s, last] [300s, last] [3600s, last] [86400s, last] |
| K8sNodeIpLimitTotal | GlobalRouter 模式集群中节点可分配的 IP 数 | – | 必填维度:tke_cluster_instance_id 非必选:node_role、node、un_instance_id |
[60s, last] [300s, last] [3600s, last] [86400s, last] |
| K8sNodeNodeAllocatableCpuForCluster | worker 节点可分配 cpu | – | 必填维度:tke_cluster_instance_id 非必选:node_role、node、un_instance_id |
[60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeNodeAllocatableMemoryForCluster | worker 节点可分配内存 | MBytes | 必填维度:tke_cluster_instance_id 非必选:node_role、node、un_instance_id |
[60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeNonVnodeAllocatablePodsForCluster | worker 节点可分配 pod 数量 | – | 必填维度:tke_cluster_instance_id 非必选:node_role、node、un_instance_id |
[60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
| K8sNodeStatusReadyForCluster | 用于统计集群总节点数 | – | 必填维度:tke_cluster_instance_id 非必选:node_role、node、un_instance_id |
[60s, avg] [300s, avg] [3600s, avg] [86400s, avg] |
三、统计粒度与时间跨度
不同统计粒度支持的时间跨度不一致,拉取监控数据时需注意时间跨度限制,具体说明如下:
| 统计粒度 | 支持最大时间跨度(结束时间-起始时间) |
| 60s | 12小时 |
| 300s | 3天 |
| 3600s | 30天 |
| 86400s | 186天 |
四、各维度对应参数总览
| 参数名称 | 类型 | 维度名称 | 维度解释 | 格式 |
| Conditions.N.Key | 必选 | tke_cluster_instance_id | 集群的维度名称 | 输入 String 类型维度名称:tke_cluster_instance_id |
| | Conditions.N.Value | tke_cluster_instance_id | 具体集群 ID | 输入具体集群 ID,例如:cls-fvkxp123 |
| Conditions.N.Key | 非必选(可不选,也可选择一项或多项) | node_role | 集群的维度名称 | 输入 String 类型维度名称:node_role |
| | Conditions.N.Value | node_role | 具体节点角色 | 输入具体节点角色,例如:node |
| | Conditions.N.Key | node | 节点名称的维度名称 | 输入 String 类型维度名称:node |
| | Conditions.N.Value | node | 具体节点名称 | 输入具体节点名称,例如:node |
| | Conditions.N.Key | un_instance_id | 节点 ID 的维度名称 | 输入 String 类型维度名称:un_instance_id |
| | Conditions.N.Value | un_instance_id | 具体节点 ID | 输入具体节点 ID,例如:ins-nwjhh123 |
五、入参说明
集群入参取值如下:
&Namespace=QCE/TKE2
&Conditions.0.Key=tke_cluster_instance_id
&Conditions.0.Value=cls-fvkxp123
根据节点角色入参取值如下:
&Namespace=QCE/TKE2
&Conditions.0.Key=tke_cluster_instance_id
&Conditions.0.Value=cls-fvkxp123
&Conditions.1.Key=node_role
&Conditions.1.Value=node
根据节点 ID 入参取值如下:
&Namespace=QCE/TKE2
&Conditions.0.Key=tke_cluster_instance_id
&Conditions.0.Value=cls-fvkxp123
&Conditions.1.Key=un_instance_id
&Conditions.1.Value=ins-nwjhh123
根据节点名称入参取值如下:
&Namespace=QCE/TKE2
&Conditions.0.Key=tke_cluster_instance_id
&Conditions.0.Value=cls-fvkxp123
&Conditions.1.Key=node
&Conditions.1.Value=node