用于 Azure Monitor 中 Prometheus 指标的最小引入配置文件
默认情况下，Azure Monitor 指标加载项会收集 Prometheus 指标的数量。 Minimal ingestion profile 是一种有助于减少指标引入量的设置，因为只有默认仪表板使用的指标，才会收集默认记录规则和默认警报。 本文介绍如何配置此设置。 本文还列出了启用 minimal ingestion profile 时默认收集的指标。 可以修改集合以启用收集更多指标，如下所示。
注意
对于基于加载项的集合，Minimal ingestion profile 设置默认处于启用状态。
以下目标默认处于“已启用/开启”状态，这意味着你不必提供任何抓取作业配置来抓取这些目标，因为指标加载项默认会自动抓取这些目标
cadvisor（job=cadvisor）nodeexporter（job=node）kubelet（job=kubelet）kube-state-metrics（job=kube-state-metrics）controlplane-apiserver（job=controlplane-apiserver）controlplane-etcd（job=controlplane-etcd）
以下目标可进行抓取，但默认情况下抓取未启用（已禁用/关闭），这意味着你不必提供任何抓取作业配置来抓取这些目标，但它们默认处于禁用/关闭状态，你需要使用 default-scrape-settings-enabled 部分下的 ama-metrics-settings-configmap 为这些目标开启/启用抓取
core-dns（job=kube-dns）kube-proxy（job=kube-proxy）api-server（job=kube-apiserver）controlplane-cluster-autoscaler（job=controlplane-cluster-autoscaler）controlplane-kube-scheduler（job=controlplane-kube-scheduler）controlplane-kube-controller-manager（job=controlplane-kube-controller-manager）
注意
所有默认目标和抓取的默认抓取频率为 30 seconds。 可以使用 default-targets-scrape-interval-settings 部分下的 ama-metrics-settings-configmap 替代每个目标。
控制平面目标具有固定的抓取间隔 30 seconds，且无法覆盖。
可在此处详细了解指标加载项使用的四种不同 configmap
配置设置
默认情况下，该设置 default-targets-metrics-keep-list.minimalIngestionProfile="true" 在指标加载项上处于启用状态。 可以在 default-targets-metrics-keep-list 部分下的 ama-metrics-settings-configmap 中指定此设置。
方案
在以下四种情况下，你可能希望自定义此行为:
每个默认目标仅引入最少指标。
这是设置 default-targets-metrics-keep-list.minimalIngestionProfile="true" 的默认行为。 对于每个默认目标，只引入下面列出的指标。
除了最少指标外，还为一个或多个默认目标引入一些其他指标。
保留 minimalIngestionProfile="true" 并指定特定于目标的相应 keeplistRegexes.* 对象，例如 keeplistRegexes.coreDns="X``Y"。 X,Y 与目标的默认指标列表合并，然后被引入。 ``
仅为目标引入一组特定的指标，而不引入其他指标。
设置 minimalIngestionProfile="false"，并指定特定于 ama-metrics-settings-configmap 中的目标的相应 default-targets-metrics-keep-list.="X``Y"。
引入为默认目标抓取的所有指标。
设置 minimalIngestionProfile="false" 并且不为该目标指定任何 default-targets-metrics-keep-list.<targetname>。 更改为 false 可能会使每个目标的指标引入量增加一个系数。
注意
up 指标不是允许/保留列表的一部分，不管指定的 keepLists 是什么，都会按每个抓取、每个目标引入指标。 此指标实际上不是抓取的，而是由指标加载项的抓取操作而生成的。 对于直方图和摘要，每个序列必须显式包含在列表中（*bucket、*sum 和 *count 序列）。
默认 ON 目标的最少引入
默认 ON 目标允许使用 minimalingestionprofile=true 列出以下指标。 默认情况下会收集以下指标，因为这些目标默认已抓取。
kubelet
kubelet_volume_stats_used_byteskubelet_node_namekubelet_running_podskubelet_running_pod_countkubelet_running_containerskubelet_running_container_countvolume_manager_total_volumeskubelet_node_config_errorkubelet_runtime_operations_totalkubelet_runtime_operations_errors_totalkubelet_runtime_operations_duration_secondskubelet_runtime_operations_duration_seconds_bucketkubelet_runtime_operations_duration_seconds_sumkubelet_runtime_operations_duration_seconds_countkubelet_pod_start_duration_secondskubelet_pod_start_duration_seconds_bucketkubelet_pod_start_duration_seconds_sumkubelet_pod_start_duration_seconds_countkubelet_pod_worker_duration_secondskubelet_pod_worker_duration_seconds_bucketkubelet_pod_worker_duration_seconds_sumkubelet_pod_worker_duration_seconds_countstorage_operation_duration_secondsstorage_operation_duration_seconds_bucketstorage_operation_duration_seconds_sumstorage_operation_duration_seconds_countstorage_operation_errors_totalkubelet_cgroup_manager_duration_secondskubelet_cgroup_manager_duration_seconds_bucketkubelet_cgroup_manager_duration_seconds_sumkubelet_cgroup_manager_duration_seconds_countkubelet_pleg_relist_duration_secondskubelet_pleg_relist_duration_seconds_bucketkubelet_pleg_relist_duration_sumkubelet_pleg_relist_duration_seconds_countkubelet_pleg_relist_interval_secondskubelet_pleg_relist_interval_seconds_bucketkubelet_pleg_relist_interval_seconds_sumkubelet_pleg_relist_interval_seconds_countrest_client_requests_totalrest_client_request_duration_secondsrest_client_request_duration_seconds_bucketrest_client_request_duration_seconds_sumrest_client_request_duration_seconds_countprocess_resident_memory_bytesprocess_cpu_seconds_totalgo_goroutineskubelet_volume_stats_capacity_byteskubelet_volume_stats_available_byteskubelet_volume_stats_inodes_usedkubelet_volume_stats_inodeskubernetes_build_info"
cadvisor
container_spec_cpu_periodcontainer_spec_cpu_quotacontainer_cpu_usage_seconds_totalcontainer_memory_rsscontainer_network_receive_bytes_totalcontainer_network_transmit_bytes_totalcontainer_network_receive_packets_totalcontainer_network_transmit_packets_totalcontainer_network_receive_packets_dropped_totalcontainer_network_transmit_packets_dropped_totalcontainer_fs_reads_totalcontainer_fs_writes_totalcontainer_fs_reads_bytes_totalcontainer_fs_writes_bytes_totalcontainer_memory_working_set_bytescontainer_memory_cachecontainer_memory_swapcontainer_cpu_cfs_throttled_periods_totalcontainer_cpu_cfs_periods_totalcontainer_memory_usage_byteskubernetes_build_info"
kube-state-metrics
kube_node_status_capacitykube_job_status_succeededkube_job_spec_completionskube_daemonset_status_desired_number_scheduledkube_daemonset_status_number_readykube_deployment_spec_replicaskube_deployment_status_replicas_readykube_pod_container_status_last_terminated_reasonkube_node_status_conditionkube_pod_container_status_restarts_totalkube_pod_container_resource_requestskube_pod_status_phasekube_pod_container_resource_limitskube_node_status_allocatablekube_pod_infokube_pod_ownerkube_resourcequotakube_statefulset_replicaskube_statefulset_status_replicaskube_statefulset_status_replicas_readykube_statefulset_status_replicas_currentkube_statefulset_status_replicas_updatedkube_namespace_status_phasekube_node_infokube_statefulset_metadata_generationkube_pod_labelskube_pod_annotationskube_horizontalpodautoscaler_status_current_replicaskube_horizontalpodautoscaler_status_desired_replicaskube_horizontalpodautoscaler_spec_min_replicaskube_horizontalpodautoscaler_spec_max_replicaskube_node_status_conditionkube_node_spec_taintkube_pod_container_status_waiting_reasonkube_job_failedkube_job_status_start_timekube_deployment_spec_replicaskube_deployment_status_replicas_availablekube_deployment_status_replicas_updatedkube_job_status_activekubernetes_build_infokube_pod_container_infokube_replicaset_ownerkube_resource_labels(ex - kube_pod_labels, kube_deployment_labels)kube_resource_annotations(ex - kube_pod_annotations, kube_deployment_annotations)
node-exporter (linux)
node_cpu_seconds_totalnode_memory_MemAvailable_bytesnode_memory_Buffers_bytesnode_memory_Cached_bytesnode_memory_MemFree_bytesnode_memory_Slab_bytesnode_memory_MemTotal_bytesnode_netstat_Tcp_RetransSegsnode_netstat_Tcp_OutSegsnode_netstat_TcpExt_TCPSynRetransnode_load1``node_load5node_load15node_disk_read_bytes_totalnode_disk_written_bytes_totalnode_disk_io_time_seconds_totalnode_filesystem_size_bytesnode_filesystem_avail_bytesnode_filesystem_readonlynode_network_receive_bytes_totalnode_network_transmit_bytes_totalnode_vmstat_pgmajfaultnode_network_receive_drop_totalnode_network_transmit_drop_totalnode_disk_io_time_weighted_seconds_totalnode_exporter_build_infonode_time_secondsnode_uname_info"
controlplane-apiserver
apiserver_request_totalapiserver_cache_list_fetched_objects_totalapiserver_cache_list_returned_objects_totalapiserver_flowcontrol_demand_seats_averageapiserver_flowcontrol_current_limit_seatsapiserver_request_sli_duration_seconds_bucketapiserver_request_sli_duration_seconds_countapiserver_request_sli_duration_seconds_sumprocess_start_time_secondsapiserver_request_duration_seconds_bucketapiserver_request_duration_seconds_countapiserver_request_duration_seconds_sumapiserver_storage_list_fetched_objects_totalapiserver_storage_list_returned_objects_totalapiserver_current_inflight_requests
controlplane-etcd
etcd_server_has_leaderrest_client_requests_totaletcd_mvcc_db_total_size_in_bytesetcd_mvcc_db_total_size_in_use_in_bytesetcd_server_slow_read_indexes_totaletcd_server_slow_apply_totaletcd_network_client_grpc_sent_bytes_totaletcd_server_heartbeat_send_failures_total
默认 OFF 目标的最少引入
默认 OFF 目标允许使用 minimalingestionprofile=true 列出以下指标。 默认情况下不会收集这些指标，因为这些目标默认不会抓取（默认为 OFF）。 可以使用 default-scrape-settings-enabled 部分下的 ama-metrics-settings-configmap 使用 default-scrape-settings-enabled.<target-name>=true 开启这些目标的抓取。
core-dns
coredns_build_infocoredns_panics_totalcoredns_dns_responses_totalcoredns_forward_responses_totalcoredns_dns_request_duration_secondscoredns_dns_request_duration_seconds_bucketcoredns_dns_request_duration_seconds_sumcoredns_dns_request_duration_seconds_countcoredns_forward_request_duration_secondscoredns_forward_request_duration_seconds_bucketcoredns_forward_request_duration_seconds_sumcoredns_forward_request_duration_seconds_countcoredns_dns_requests_totalcoredns_forward_requests_totalcoredns_cache_hits_totalcoredns_cache_misses_totalcoredns_cache_entriescoredns_plugin_enabledcoredns_dns_request_size_bytescoredns_dns_request_size_bytes_bucketcoredns_dns_request_size_bytes_sumcoredns_dns_request_size_bytes_countcoredns_dns_response_size_bytescoredns_dns_response_size_bytes_bucketcoredns_dns_response_size_bytes_sumcoredns_dns_response_size_bytes_countcoredns_dns_response_size_bytescoredns_dns_response_size_bytes_bucketcoredns_dns_response_size_bytes_sumcoredns_dns_response_size_bytes_countprocess_resident_memory_bytesprocess_cpu_seconds_totalgo_goroutineskubernetes_build_info"
kube-proxy
kubeproxy_sync_proxy_rules_duration_secondskubeproxy_sync_proxy_rules_duration_seconds_bucketkubeproxy_sync_proxy_rules_duration_seconds_sumkubeproxy_sync_proxy_rules_duration_seconds_countkubeproxy_network_programming_duration_secondskubeproxy_network_programming_duration_secondskubeproxy_network_programming_duration_seconds_bucketkubeproxy_network_programming_duration_seconds_sumkubeproxy_network_programming_duration_seconds_countrest_client_requests_totalrest_client_request_duration_secondsrest_client_request_duration_seconds_bucketrest_client_request_duration_seconds_sumrest_client_request_duration_seconds_countprocess_resident_memory_bytesprocess_cpu_seconds_totalgo_goroutineskubernetes_build_info"
api-server
apiserver_request_duration_secondsapiserver_request_duration_seconds_bucketapiserver_request_duration_seconds_sumapiserver_request_duration_seconds_countapiserver_request_totalworkqueue_adds_total``workqueue_depthworkqueue_queue_duration_secondsworkqueue_queue_duration_seconds_bucketworkqueue_queue_duration_seconds_sumworkqueue_queue_duration_seconds_countprocess_resident_memory_bytesprocess_cpu_seconds_totalgo_goroutineskubernetes_build_info"
windows-exporter (job=windows-exporter)
windows_system_system_up_timewindows_cpu_time_totalwindows_memory_available_byteswindows_os_visible_memory_byteswindows_memory_cache_byteswindows_memory_modified_page_list_byteswindows_memory_standby_cache_core_byteswindows_memory_standby_cache_normal_priority_byteswindows_memory_standby_cache_reserve_byteswindows_memory_swap_page_operations_totalwindows_logical_disk_read_seconds_totalwindows_logical_disk_write_seconds_totalwindows_logical_disk_size_byteswindows_logical_disk_free_byteswindows_net_bytes_totalwindows_net_packets_received_discarded_totalwindows_net_packets_outbound_discarded_totalwindows_container_availablewindows_container_cpu_usage_seconds_totalwindows_container_memory_usage_commit_byteswindows_container_memory_usage_private_working_set_byteswindows_container_network_receive_bytes_totalwindows_container_network_transmit_bytes_total
kube-proxy-windows (job=kube-proxy-windows)
kubeproxy_sync_proxy_rules_duration_secondskubeproxy_sync_proxy_rules_duration_seconds_bucketkubeproxy_sync_proxy_rules_duration_seconds_sumkubeproxy_sync_proxy_rules_duration_seconds_countrest_client_requests_totalrest_client_request_duration_secondsrest_client_request_duration_seconds_bucketrest_client_request_duration_seconds_sumrest_client_request_duration_seconds_countprocess_resident_memory_bytesprocess_cpu_seconds_totalgo_goroutines
controlplane-cluster-autoscaler
rest_client_requests_totalcluster_autoscaler_last_activitycluster_autoscaler_cluster_safe_to_autoscalecluster_autoscaler_scale_down_in_cooldowncluster_autoscaler_scaled_up_nodes_totalcluster_autoscaler_unneeded_nodes_countcluster_autoscaler_unschedulable_pods_countcluster_autoscaler_nodes_countcloudprovider_azure_api_request_errorscloudprovider_azure_api_request_duration_seconds_bucketcloudprovider_azure_api_request_duration_seconds_count
controlplane-kube-scheduler
scheduler_pending_podsscheduler_unschedulable_podsscheduler_pod_scheduling_attemptsscheduler_queue_incoming_pods_totalscheduler_preemption_attempts_totalscheduler_preemption_victimsscheduler_scheduling_attempt_duration_secondsscheduler_schedule_attempts_totalscheduler_pod_scheduling_duration_seconds
controlplane-kube-controller-manager
rest_client_request_duration_secondsrest_client_requests_totalworkqueue_depth