容器监控实践，从入门到放弃

相关文档

prometheus 入门到放弃

https://yunlzheng.gitbook.io/prometheus-book/

prometheus关于容器监控主要参数和指标

https://github.com/google/cadvisor/blob/master/docs/storage/prometheus.md

prometheus node_exporter下载地址

https://github.com/prometheus/node_exporter/releases/

kubernetes监控指标架构图

https://github.com/kubernetes/community/blob/master/contributors/design-proposals/instrumentation/monitoring_architecture.md

kubernetes/kube-state-metrics 项目GitHub地址

https://github.com/kubernetes/kube-state-metrics#metrics-documentation

kube-state-metrics K8S监控指标全部文档

https://github.com/kubernetes/kube-state-metrics/tree/master/docs

针对基础架构，微服务和应用程序的实时云监视和可观察性

https://docs.signalfx.com/en/latest/integrations/agent/monitors/kubernetes-proxy.html

阿里云 K8S监控报警全部指标
https://help.aliyun.com/document_detail/176180.html?spm=a2c4g.11186623.6.659.598c2d39N3EVnR

报警名称	表达式	采集数据时间（分钟）	报警触发条件
PodCpu75	100 * (sum(rate(container_cpu_usage_seconds_total[1m])) by (pod_name) / sum(label_replace(kube_pod_container_resource_limits_cpu_cores, "pod_name", "$1", "pod", "(.*)")) by (pod_name))>75	7	Pod的CPU使用率大于75%。
PodMemory75	100 * (sum(container_memory_working_set_bytes) by (pod_name) / sum(label_replace(kube_pod_container_resource_limits_memory_bytes, "pod_name", "$1", "pod", "(.*)")) by (pod_name))>75	5	Pod的内存使用率大于75%。
pod_status_no_running	sum (kube_pod_status_phase{phase!="Running"}) by (pod,phase)	5	Pod的状态为未运行。
PodMem4GbRestart	(sum (container_memory_working_set_bytes{id!="/"})by (pod_name,container_name) /1024/1024/1024)>4	5	Pod的内存大于4GB。
PodRestart	sum (increase (kube_pod_container_status_restarts_total{}[2m])) by (namespace,pod) >0	5	Pod重启。

K8s报警规则

报警名称	表达式	采集数据时间（分钟）	报警触发条件
KubeStateMetricsListErrors	(sum(rate(kube_state_metrics_list_total{job="kube-state-metrics",result="error"}[5m])) / sum(rate(kube_state_metrics_list_total{job="kube-state-metrics"}[5m]))) > 0.01	15	Metric List出错。
KubeStateMetricsWatchErrors	(sum(rate(kube_state_metrics_watch_total{job="kube-state-metrics",result="error"}[5m])) / sum(rate(kube_state_metrics_watch_total{job="kube-state-metrics"}[5m]))) > 0.01	15	Metric Watch出错。
NodeFilesystemAlmostOutOfSpace	( node_filesystem_avail_bytes{job="node-exporter",fstype!=""} / node_filesystem_size_bytes{job="node-exporter",fstype!=""} * 100 < 5 and node_filesystem_readonly{job="node-exporter",fstype!=""} == 0 )	60	Node文件系统即将无空间。
NodeFilesystemSpaceFillingUp	( node_filesystem_avail_bytes{job="node-exporter",fstype!=""} / node_filesystem_size_bytes{job="node-exporter",fstype!=""} * 100 < 40 and predict_linear(node_filesystem_avail_bytes{job="node-exporter",fstype!=""}[6h], 24*60*60) < 0 and node_filesystem_readonly{job="node-exporter",fstype!=""} == 0 )	60	Node文件系统空间即将占满。
NodeFilesystemFilesFillingUp	( node_filesystem_files_free{job="node-exporter",fstype!=""} / node_filesystem_files{job="node-exporter",fstype!=""} * 100 < 40 and predict_linear(node_filesystem_files_free{job="node-exporter",fstype!=""}[6h], 24*60*60) < 0 and node_filesystem_readonly{job="node-exporter",fstype!=""} == 0 )	60	Node文件系统文件即将占满。
NodeFilesystemAlmostOutOfFiles	( node_filesystem_files_free{job="node-exporter",fstype!=""} / node_filesystem_files{job="node-exporter",fstype!=""} * 100 < 3 and node_filesystem_readonly{job="node-exporter",fstype!=""} == 0 )	60	Node文件系统几乎无文件。
NodeNetworkReceiveErrs	increase(node_network_receive_errs_total[2m]) > 10	60	Node网络接收错误。
NodeNetworkTransmitErrs	increase(node_network_transmit_errs_total[2m]) > 10	60	Node网络传输错误。
NodeHighNumberConntrackEntriesUsed	(node_nf_conntrack_entries / node_nf_conntrack_entries_limit) > 0.75	无	使用大量Conntrack条目。
NodeClockSkewDetected	( node_timex_offset_seconds > 0.05 and deriv(node_timex_offset_seconds[5m]) >= 0 ) or ( node_timex_offset_seconds < -0.05 and deriv(node_timex_offset_seconds[5m]) <= 0 )	10	出现时间偏差。
NodeClockNotSynchronising	min_over_time(node_timex_sync_status[5m]) == 0	10	出现时间不同步。
KubePodCrashLooping	rate(kube_pod_container_status_restarts_total{job="kube-state-metrics"}[15m]) * 60 * 5 > 0	15	出现循环崩溃。
KubePodNotReady	sum by (namespace, pod) (max by(namespace, pod) (kube_pod_status_phase{job="kube-state-metrics", phase=~"Pending|Unknown"}) * on(namespace, pod) group_left(owner_kind) max by(namespace, pod, owner_kind) (kube_pod_owner{owner_kind!="Job"})) > 0	15	Pod未准备好。
KubeDeploymentGenerationMismatch	kube_deployment_status_observed_generation{job="kube-state-metrics"} != kube_deployment_metadata_generation{job="kube-state-metrics"}	15	出现部署版本不匹配。
KubeDeploymentReplicasMismatch	( kube_deployment_spec_replicas{job="kube-state-metrics"} != kube_deployment_status_replicas_available{job="kube-state-metrics"} ) and ( changes(kube_deployment_status_replicas_updated{job="kube-state-metrics"}[5m]) == 0 )	15	出现部署副本不匹配。
KubeStatefulSetReplicasMismatch	( kube_statefulset_status_replicas_ready{job="kube-state-metrics"} != kube_statefulset_status_replicas{job="kube-state-metrics"} ) and ( changes(kube_statefulset_status_replicas_updated{job="kube-state-metrics"}[5m]) == 0 )	15	状态集副本不匹配。
KubeStatefulSetGenerationMismatch	kube_statefulset_status_observed_generation{job="kube-state-metrics"} != kube_statefulset_metadata_generation{job="kube-state-metrics"}	15	状态集版本不匹配。
KubeStatefulSetUpdateNotRolledOut	max without (revision) ( kube_statefulset_status_current_revision{job="kube-state-metrics"} unless kube_statefulset_status_update_revision{job="kube-state-metrics"} ) * ( kube_statefulset_replicas{job="kube-state-metrics"} != kube_statefulset_status_replicas_updated{job="kube-state-metrics"} )	15	状态集更新未推出。
KubeDaemonSetRolloutStuck	kube_daemonset_status_number_ready{job="kube-state-metrics"} / kube_daemonset_status_desired_number_scheduled{job="kube-state-metrics"} < 1.00	15	DaemonSet推出回退。
KubeContainerWaiting	sum by (namespace, pod, container) (kube_pod_container_status_waiting_reason{job="kube-state-metrics"}) > 0	60	容器等待。
KubeDaemonSetNotScheduled	kube_daemonset_status_desired_number_scheduled{job="kube-state-metrics"} - kube_daemonset_status_current_number_scheduled{job="kube-state-metrics"} > 0	10	DaemonSet无计划。
KubeDaemonSetMisScheduled	kube_daemonset_status_number_misscheduled{job="kube-state-metrics"} > 0	15	Daemon缺失计划。
KubeCronJobRunning	time() - kube_cronjob_next_schedule_time{job="kube-state-metrics"} > 3600	60	若Cron任务完成时间大于1小。
KubeJobCompletion	kube_job_spec_completions{job="kube-state-metrics"} - kube_job_status_succeeded{job="kube-state-metrics"} > 0	60	任务完成。
KubeJobFailed	kube_job_failed{job="kube-state-metrics"} > 0	15	任务失败。
KubeHpaReplicasMismatch	(kube_hpa_status_desired_replicas{job="kube-state-metrics"} != kube_hpa_status_current_replicas{job="kube-state-metrics"}) and changes(kube_hpa_status_current_replicas[15m]) == 0	15	HPA副本不匹配。
KubeHpaMaxedOut	kube_hpa_status_current_replicas{job="kube-state-metrics"} == kube_hpa_spec_max_replicas{job="kube-state-metrics"}	15	HPA副本超过最大值。
KubeCPUOvercommit	sum(namespace:kube_pod_container_resource_requests_cpu_cores:sum{}) / sum(kube_node_status_allocatable_cpu_cores) > (count(kube_node_status_allocatable_cpu_cores)-1) / count(kube_node_status_allocatable_cpu_cores)	5	CPU过载。
KubeMemoryOvercommit	sum(namespace:kube_pod_container_resource_requests_memory_bytes:sum{}) / sum(kube_node_status_allocatable_memory_bytes) > (count(kube_node_status_allocatable_memory_bytes)-1) / count(kube_node_status_allocatable_memory_bytes)	5	存储过载。
KubeCPUQuotaOvercommit	sum(kube_resourcequota{job="kube-state-metrics", type="hard", resource="cpu"}) / sum(kube_node_status_allocatable_cpu_cores) > 1.5	5	CPU额度过载。
KubeMemoryQuotaOvercommit	sum(kube_resourcequota{job="kube-state-metrics", type="hard", resource="memory"}) / sum(kube_node_status_allocatable_memory_bytes{job="node-exporter"}) > 1.5	5	存储额度过载。
KubeQuotaExceeded	kube_resourcequota{job="kube-state-metrics", type="used"} / ignoring(instance, job, type) (kube_resourcequota{job="kube-state-metrics", type="hard"} > 0) > 0.90	15	若配额超过限制。
CPUThrottlingHigh	sum(increase(container_cpu_cfs_throttled_periods_total{container!="", }[5m])) by (container, pod, namespace) / sum(increase(container_cpu_cfs_periods_total{}[5m])) by (container, pod, namespace) > ( 25 / 100 )	15	CPU过热。
KubePersistentVolumeFillingUp	kubelet_volume_stats_available_bytes{job="kubelet", metrics_path="/metrics"} / kubelet_volume_stats_capacity_bytes{job="kubelet", metrics_path="/metrics"} < 0.03	1	存储卷容量即将不足。
KubePersistentVolumeErrors	kube_persistentvolume_status_phase{phase=~"Failed|Pending",job="kube-state-metrics"} > 0	5	存储卷容量出错。
KubeVersionMismatch	count(count by (gitVersion) (label_replace(kubernetes_build_info{job!~"kube-dns|coredns"},"gitVersion","$1","gitVersion","(v[0-9]*.[0-9]*.[0-9]*).*"))) > 1	15	版本不匹配。
KubeClientErrors	(sum(rate(rest_client_requests_total{code=~"5.."}[5m])) by (instance, job) / sum(rate(rest_client_requests_total[5m])) by (instance, job)) > 0.01	15	客户端出错。
KubeAPIErrorBudgetBurn	sum(apiserver_request:burnrate1h) > (14.40 * 0.01000) and sum(apiserver_request:burnrate5m) > (14.40 * 0.01000)	2	API错误过多。
KubeAPILatencyHigh	( cluster:apiserver_request_duration_seconds:mean5m{job="apiserver"} > on (verb) group_left() ( avg by (verb) (cluster:apiserver_request_duration_seconds:mean5m{job="apiserver"} >= 0) + 2*stddev by (verb) (cluster:apiserver_request_duration_seconds:mean5m{job="apiserver"} >= 0) ) ) > on (verb) group_left() 1.2 * avg by (verb) (cluster:apiserver_request_duration_seconds:mean5m{job="apiserver"} >= 0) and on (verb,resource) cluster_quantile:apiserver_request_duration_seconds:histogram_quantile{job="apiserver",quantile="0.99"} > 1	5	API延迟过高。
KubeAPIErrorsHigh	sum(rate(apiserver_request_total{job="apiserver",code=~"5.."}[5m])) by (resource,subresource,verb) / sum(rate(apiserver_request_total{job="apiserver"}[5m])) by (resource,subresource,verb) > 0.05	10	API错误过多。
KubeClientCertificateExpiration	apiserver_client_certificate_expiration_seconds_count{job="apiserver"} > 0 and on(job) histogram_quantile(0.01, sum by (job, le) (rate(apiserver_client_certificate_expiration_seconds_bucket{job="apiserver"}[5m]))) < 604800	无	客户端认证过期。
AggregatedAPIErrors	sum by(name, namespace)(increase(aggregator_unavailable_apiservice_count[5m])) > 2	无	聚合API出错。
AggregatedAPIDown	sum by(name, namespace)(sum_over_time(aggregator_unavailable_apiservice[5m])) > 0	5	聚合API下线。
KubeAPIDown	absent(up{job="apiserver"} == 1)	15	API下线。
KubeNodeNotReady	kube_node_status_condition{job="kube-state-metrics",condition="Ready",status="true"} == 0	15	Node未准备好。
KubeNodeUnreachable	kube_node_spec_taint{job="kube-state-metrics",key="node.kubernetes.io/unreachable",effect="NoSchedule"} == 1	2	Node无法获取。
KubeletTooManyPods	max(max(kubelet_running_pod_count{job="kubelet", metrics_path="/metrics"}) by(instance) * on(instance) group_left(node) kubelet_node_name{job="kubelet", metrics_path="/metrics"}) by(node) / max(kube_node_status_capacity_pods{job="kube-state-metrics"} != 1) by(node) > 0.95	15	Pod过多。
KubeNodeReadinessFlapping	sum(changes(kube_node_status_condition{status="true",condition="Ready"}[15m])) by (node) > 2	15	准备状态变更次数过多。
KubeletPlegDurationHigh	node_quantile:kubelet_pleg_relist_duration_seconds:histogram_quantile{quantile="0.99"} >= 10	5	PLEG持续时间过长。
KubeletPodStartUpLatencyHigh	histogram_quantile(0.99, sum(rate(kubelet_pod_worker_duration_seconds_bucket{job="kubelet", metrics_path="/metrics"}[5m])) by (instance, le)) * on(instance) group_left(node) kubelet_node_name{job="kubelet", metrics_path="/metrics"} > 60	15	Pod启动延迟过高。
KubeletDown	absent(up{job="kubelet", metrics_path="/metrics"} == 1)	15	Kubelet下线。
KubeSchedulerDown	absent(up{job="kube-scheduler"} == 1)	15	Kubelet日程下线。
KubeControllerManagerDown	absent(up{job="kube-controller-manager"} == 1)	15	Controller Manager下线。
TargetDown	100 * (count(up == 0) BY (job, namespace, service) / count(up) BY (job, namespace, service)) > 10	10	目标下线。
NodeNetworkInterfaceFlapping	changes(node_network_up{job="node-exporter",device!~"veth.+"}[2m]) > 2	2	网络接口状态变更过频繁。