Node Exporter

grafana-agent 内置了 node_exporter, 可以通过在配置文件中 integrations 部分定义 node_exporter_config 来开启该功能。

配置并启用node_exporter

下面是开启了node_exporter的配置文件示例，生成的配置文件保存为 ./grafana-agent-cfg.yaml:

  1. cat <<EOF > ./grafana-agent-cfg.yaml
  2. # grafana-agent 本身的配置
  3. server:
  4.   log_level: info
  5.   http_listen_port: 12345
  7. # grafana-agent 抓取 metrics 的相关配置（类似于prometheus的scrape_configs）
  8. metrics:
  9.   global:
  10.     scrape_interval: 15s
  11.     scrape_timeout: 10s
  12.     remote_write:
  13.       - url: https://n9e-server:19000/prometheus/v1/write
  14.         basic_auth:
  15.           username: <string>
  16.           password: <string>
  18. integrations:
  19.   node_exporter:
  20.     enabled: true
  21. EOF

注意： remote_write 可以配置在 global 部分，也可以针对每个 integration 单独配置不同的remote_write 地址。

重启grafana-agent后，通过以下两个命令，验证 node_exporter 工作是否符合预期。

curl http://localhost:12345/integrations/node_exporter/metrics ，预期输出如下内容：

  1. node_boot_time_seconds 1.643256088e+09
  2. node_context_switches_total 1.5136425575e+10
  3. node_cooling_device_cur_state{name="0",type="Processor"} 0
  4. node_cooling_device_cur_state{name="1",type="Processor"} 0
  5. node_cooling_device_cur_state{name="2",type="Processor"} 0
  6. node_cooling_device_cur_state{name="3",type="Processor"} 0
  7. node_cooling_device_max_state{name="0",type="Processor"} 0
  8. node_cooling_device_max_state{name="1",type="Processor"} 0
  9. node_cooling_device_max_state{name="2",type="Processor"} 0
  10. node_cooling_device_max_state{name="3",type="Processor"} 0
  11. node_cpu_seconds_total{cpu="0",mode="idle"} 1.66906519e+06
  12. node_cpu_seconds_total{cpu="0",mode="iowait"} 5031.48
  13. node_cpu_seconds_total{cpu="0",mode="irq"} 0
  14. node_cpu_seconds_total{cpu="0",mode="nice"} 82.84
  15. node_cpu_seconds_total{cpu="0",mode="softirq"} 2332.39

curl http://localhost:12345/agent/api/v1/targets | jq，预期输出如下内容：

  1. {
  2.   "status": "success",
  3.   "data": [
  4.     {
  5.       "instance": "b81030837ec7f1d162489cb4009325c9",
  6.       "target_group": "integrations/node_exporter",
  7.       "endpoint": "http://127.0.0.1:12345/integrations/node_exporter/metrics",
  8.       "state": "up",
  9.       "labels": {
  10.         "agent_hostname": "tt-fc-dev01.nj",
  11.         "instance": "tt-fc-dev01.nj:12345",
  12.         "job": "integrations/node_exporter"
  13.       },
  14.       "discovered_labels": {
  15.         "__address__": "127.0.0.1:12345",
  16.         "__metrics_path__": "/integrations/node_exporter/metrics",
  17.         "__scheme__": "http",
  18.         "__scrape_interval__": "15s",
  19.         "__scrape_timeout__": "10s",
  20.         "agent_hostname": "tt-fc-dev01.nj",
  21.         "job": "integrations/node_exporter"
  22.       },
  23.       "last_scrape": "2022-02-16T18:53:08.79288957+08:00",
  24.       "scrape_duration_ms": 20,
  25.       "scrape_error": ""
  26.     },
  27.     {
  28.       "instance": "b81030837ec7f1d162489cb4009325c9",
  29.       "target_group": "local_scrape",
  30.       "endpoint": "http://127.0.0.1:12345/metrics",
  31.       "state": "up",
  32.       "labels": {
  33.         "cluster": "txnjdev01",
  34.         "instance": "127.0.0.1:12345",
  35.         "job": "local_scrape"
  36.       },
  37.       "discovered_labels": {
  38.         "__address__": "127.0.0.1:12345",
  39.         "__metrics_path__": "/metrics",
  40.         "__scheme__": "http",
  41.         "__scrape_interval__": "15s",
  42.         "__scrape_timeout__": "10s",
  43.         "cluster": "txnjdev01",
  44.         "job": "local_scrape"
  45.       },
  46.       "last_scrape": "2022-02-16T18:53:22.336820442+08:00",
  47.       "scrape_duration_ms": 4,
  48.       "scrape_error": ""
  49.     }
  50.   ]
  51. }

可以看到，上面的返回结果的 targets 列表中，已经新增了一个instance，其 job 为 integrations/node_exporter，这说明 node_exporter 已经在正常工作了。

注意：如果 grafana-agent 是运行在容器中时，那么要做以下修改调整：

  1. 确保在运行容器时，将宿主机的相关目录映射到容器中，如下所示，即 -v "/:/host/root"、 -v "/sys:/host/sys"、-v "/proc:/host/proc".
  1. docker run \
  2.   --net="host" \
  3.   --pid="host" \
  4.   --cap-add=SYS_TIME \
  5.   -d \
  6.   -v "/:/host/root:ro" \
  7.   -v "/sys:/host/sys:ro" \
  8.   -v "/proc:/host/proc:ro" \
  9.   -v /tmp/grafana-agent:/etc/agent/data \
  10.   -v /tmp/grafana-agent-config.yaml:/etc/agent/agent.yaml \
  11.   grafana/agent:v0.23.0 \
  12.   --config.file=/etc/agent/agent.yaml \
  13.   --metrics.wal-directory=/etc/agent/data
  1. 其中，配置文件 /tmp/grafana-agent-config.yaml 中 node_exporter 部分要指定 rootfs/sysfs/procfs 在容器中的路径，您可以运行以下命令生成该测试配置文件（当然，您需要把 remote_write 替换为适合您的地址）。
  1. cat <<EOF > /tmp/grafana-agent-config.yaml
  2. server:
  3.   log_level: info
  4.   http_listen_port: 12345
  6. metrics:
  7.   global:
  8.     scrape_interval: 15s
  9.     scrape_timeout: 10s
  10.     remote_write:
  11.       - url: https://n9e-server:19000/prometheus/v1/write
  12.         basic_auth:
  13.           username: <string>
  14.           password: <string>
  16. integrations:
  17.   node_exporter:
  18.     enabled: true
  19.     rootfs_path: /host/root
  20.     sysfs_path: /host/sys
  21.     procfs_path: /host/proc
  22. EOF

注意：如果 grafana-agent 是运行在 K8s 环境中，那么调整步骤如下：

  1. 推荐将 grafana-agent 的配置文件存储在configmap中, manifest文件如下：
  1. cat <<EOF |
  2. apiVersion: v1
  3. kind: ConfigMap
  4. metadata:
  5.   name: grafana-agent
  6.   namespace: ${NAMESPACE}
  7. data:
  8.   agent.yaml: |
  9.     server:
  10.       log_level: info
  11.       http_listen_port: 12345
  13.     metrics:
  14.       global:
  15.         scrape_interval: 15s
  16.         remote_write:
  17.           - url: 'https://n9e-server:19000/prometheus/v1/write'
  18.             basic_auth:
  19.               username: ${FC_USERNAME}
  20.               password: ${FC_PASSWORD}
  22.     integrations:
  23.       agent:
  24.         enabled: true
  25.       node_exporter:
  26.         enabled: true 
  27. EOF
  29. envsubst | kubectl apply -f -
  30. kubectl describe configmap grafana-agent
  1. 生成grafana-agent的pod manifest文件如下，并创建相应Pod实例：
  1. cat << EOF |
  2. apiVersion: v1
  3. kind: Pod
  4. metadata:
  5.   name: grafana-agent
  6.   namespace: ${NAMESPACE}
  7. spec:
  8.   containers:
  9.   - image: grafana/agent:v0.23.0
  10.     name: grafana-agent
  11.     args:
  12.     - --config.file=/fcetc/agent.yaml
  13.     - --metrics.wal-directory=/etc/agent/data
  14.     securityContext:
  15.       capabilities:
  16.         add: ["SYS_TIME"]
  17.       privileged: true
  18.       runAsUser: 0
  19.     volumeMounts:
  20.     - name: rootfs
  21.       mountPath: /host/root
  22.       readOnly: true
  23.     - name: sysfs
  24.       mountPath: /host/sys
  25.       readOnly: true
  26.     - name: procfs
  27.       mountPath: /host/proc
  28.       readOnly: true
  29.     - name: fccfg
  30.       mountPath: /fcetc
  31.   hostPID: true
  32.   hostNetwork: true
  33.   dnsPolicy: ClusterFirstWithHostNet
  34.   volumes:
  35.   - name: rootfs
  36.     hostPath:
  37.       path: /
  38.   - name: sysfs
  39.     hostPath:
  40.       path: /sys
  41.   - name: procfs
  42.     hostPath:
  43.       path: /proc
  44.   - name: fccfg
  45.     configMap:
  46.       name: grafana-agent
  47. EOF
  49. envsubst |kubectl apply -f -
  50. kubectl logs grafana-agent #查看 grafana-agent 的日志

node_exporter采集的关键指标解析

  1. # SYSTEM
  2. # CPU context switch 次数
  3. node_context_switches_total: context_switches
  5. # Interrupts 次数
  6. node_intr_total: Interrupts
  8. # 运行的进程数
  9. node_procs_running: Processes in runnable state
  11. # 熵池大小
  12. node_entropy_available_bits: Entropy available to random number generators
  14. node_time_seconds: System time in seconds since epoch (1970)
  15. node_boot_time_seconds: Node boot time, in unixtime
  17. # CPU
  18. node_cpu_seconds_total: Seconds the CPUs spent in each mode
  19. node_load1: cpu load 1m
  20. node_load5: cpu load 5m
  21. node_load15: cpu load 15m
  23. # MEM
  24. # 内核态
  25. # 用户追踪已从交换区获取但尚未修改的页面的内存
  26. node_memory_SwapCached_bytes: Memory that keeps track of pages that have been fetched from swap but not yet been modified
  28. # 内核用于缓存数据结构供自己使用的内存
  29. node_memory_Slab_bytes: Memory used by the kernel to cache data structures for its own use
  31. # slab中可回收的部分
  32. node_memory_SReclaimable_bytes: SReclaimable - Part of Slab, that might be reclaimed, such as caches
  34. # slab中不可回收的部分
  35. node_memory_SUnreclaim_bytes: Part of Slab, that cannot be reclaimed on memory pressure
  37. # Vmalloc内存区的大小
  38. node_memory_VmallocTotal_bytes: Total size of vmalloc memory area
  40. # vmalloc已分配的内存，虚拟地址空间上的连续的内存
  41. node_memory_VmallocUsed_bytes: Amount of vmalloc area which is used
  43. # vmalloc区可用的连续最大快的大小，通过此指标可以知道vmalloc可分配连续内存的最大值
  44. node_memory_VmallocChunk_bytes: Largest contigious block of vmalloc area which is free
  46. # 内存的硬件故障删除掉的内存页的总大小
  47. node_memory_HardwareCorrupted_bytes: Amount of RAM that the kernel identified as corrupted / not working
  49. # 用于在虚拟和物理内存地址之间映射的内存
  50. node_memory_PageTables_bytes: Memory used to map between virtual and physical memory addresses (gauge)
  52. # 内核栈内存，常驻内存，不可回收
  53. node_memory_KernelStack_bytes: Kernel memory stack. This is not reclaimable
  55. # 用来访问高端内存，复制高端内存的临时buffer，称为“bounce buffering”，会降低I/O 性能
  56. node_memory_Bounce_bytes: Memory used for block device bounce buffers
  58. #用户态
  59. # 单个巨页大小
  60. node_memory_Hugepagesize_bytes: Huge Page size
  62. # 系统分配的常驻巨页数
  63. node_memory_HugePages_Total: Total size of the pool of huge pages
  65. # 系统空闲的巨页数
  66. node_memory_HugePages_Free: Huge pages in the pool that are not yet allocated
  68. # 进程已申请但未使用的巨页数
  69. node_memory_HugePages_Rsvd: Huge pages for which a commitment to allocate from the pool has been made, but no allocation
  71. # 超过系统设定的常驻HugePages数量的个数
  72. node_memory_HugePages_Surp: Huge pages in the pool above the value in /proc/sys/vm/nr_hugepages
  74. # 透明巨页 Transparent HugePages (THP)
  75. node_memory_AnonHugePages_bytes: Memory in anonymous huge pages
  77. # inactivelist中的File-backed内存
  78. node_memory_Inactive_file_bytes: File-backed memory on inactive LRU list
  80. # inactivelist中的Anonymous内存
  81. node_memory_Inactive_anon_bytes: Anonymous and swap cache on inactive LRU list, including tmpfs (shmem)
  83. # activelist中的File-backed内存
  84. node_memory_Active_file_bytes: File-backed memory on active LRU list
  86. # activelist中的Anonymous内存
  87. node_memory_Active_anon_bytes: Anonymous and swap cache on active least-recently-used (LRU) list, including tmpfs
  89. # 禁止换出的页，对应 Unevictable 链表
  90. node_memory_Unevictable_bytes: Amount of unevictable memory that can't be swapped out for a variety of reasons
  92. # 共享内存
  93. node_memory_Shmem_bytes: Used shared memory (shared between several processes, thus including RAM disks)
  95. # 匿名页内存大小
  96. node_memory_AnonPages_bytes: Memory in user pages not backed by files
  98. # 被关联的内存页大小
  99. node_memory_Mapped_bytes: Used memory in mapped pages files which have been mmaped, such as libraries
  101. # file-backed内存页缓存大小
  102. node_memory_Cached_bytes: Parked file data (file content) cache
  104. # 系统中有多少匿名页曾经被swap-out、现在又被swap-in并且swap-in之后页面中的内容一直没发生变化
  105. node_memory_SwapCached_bytes: Memory that keeps track of pages that have been fetched from swap but not yet been modified
  107. # 被mlock()系统调用锁定的内存大小
  108. node_memory_Mlocked_bytes: Size of pages locked to memory using the mlock() system call
  110. # 块设备(block device)所占用的缓存页
  111. node_memory_Buffers_bytes: Block device (e.g. harddisk) cache
  113. node_memory_SwapTotal_bytes: Memory information field SwapTotal_bytes
  114. node_memory_SwapFree_bytes: Memory information field SwapFree_bytes
  116. # DISK
  118. node_filesystem_files_free: Filesystem space available to non-root users in byte
  119. node_filesystem_free_bytes: Filesystem free space in bytes
  120. node_filesystem_size_bytes: Filesystem size in bytes
  122. node_filesystem_files_free: Filesystem total free file nodes
  123. node_filesystem_files: Filesystem total free file nodes
  125. node_filefd_maximum: Max open files
  126. node_filefd_allocated: Open files
  128. node_filesystem_readonly: Filesystem read-only status
  129. node_filesystem_device_error: Whether an error occurred while getting statistics for the given device
  131. node_disk_reads_completed_total: The total number of reads completed successfully
  132. node_disk_writes_completed_total: The total number of writes completed successfully
  133. node_disk_reads_merged_total: The number of reads merged
  134. node_disk_writes_merged_total: The number of writes merged
  136. node_disk_read_bytes_total: The total number of bytes read successfully
  137. node_disk_written_bytes_total: The total number of bytes written successfully
  139. node_disk_io_time_seconds_total: Total seconds spent doing I/Os
  141. node_disk_read_time_seconds_total: The total number of seconds spent by all reads
  142. node_disk_write_time_seconds_total: The total number of seconds spent by all writes
  144. node_disk_io_time_weighted_seconds_total: The weighted of seconds spent doing I/Os
  146. # NET
  147. node_network_receive_bytes_total: Network device statistic receive_bytes (counter)
  148. node_network_transmit_bytes_total: Network device statistic transmit_bytes (counter)
  150. node_network_receive_packets_total: Network device statistic receive_bytes
  151. node_network_transmit_packets_total: Network device statistic transmit_bytes
  153. node_network_receive_errs_total: Network device statistic receive_errs
  154. node_network_transmit_errs_total: Network device statistic transmit_errs
  156. node_network_receive_drop_total: Network device statistic receive_drop
  157. node_network_transmit_drop_total: Network device statistic transmit_drop
  159. node_nf_conntrack_entries: Number of currently allocated flow entries for connection tracking
  161. node_sockstat_TCP_alloc: Number of TCP sockets in state alloc
  162. node_sockstat_TCP_inuse: Number of TCP sockets in state inuse
  163. node_sockstat_TCP_orphan: Number of TCP sockets in state orphan
  164. node_sockstat_TCP_tw: Number of TCP sockets in state tw
  165. node_netstat_Tcp_CurrEstab: Statistic TcpCurrEstab
  167. node_sockstat_sockets_used: Number of IPv4 sockets in use

node_expoter integration 完整的配置项说明

  1.   # Enables the node_exporter integration, allowing the Agent to automatically
  2.   # collect system metrics from the host UNIX system.
  3.   [enabled: <boolean> | default = false]
  5.   # Sets an explicit value for the instance label when the integration is
  6.   # self-scraped. Overrides inferred values.
  7.   #
  8.   # The default value for this integration is inferred from the agent hostname
  9.   # and HTTP listen port, delimited by a colon.
  10.   [instance: <string>]
  12.   # Automatically collect metrics from this integration. If disabled,
  13.   # the node_exporter integration will be run but not scraped and thus not remote-written. Metrics for the
  14.   # integration will be exposed at /integrations/node_exporter/metrics and can
  15.   # be scraped by an external process.
  16.   [scrape_integration: <boolean> | default = <integrations_config.scrape_integrations>]
  18.   # How often should the metrics be collected? Defaults to
  19.   # prometheus.global.scrape_interval.
  20.   [scrape_interval: <duration> | default = <global_config.scrape_interval>]
  22.   # The timtout before considering the scrape a failure. Defaults to
  23.   # prometheus.global.scrape_timeout.
  24.   [scrape_timeout: <duration> | default = <global_config.scrape_timeout>]
  26.   # Allows for relabeling labels on the target.
  27.   relabel_configs:
  28.     [- <relabel_config> ... ]
  30.   # Relabel metrics coming from the integration, allowing to drop series
  31.   # from the integration that you don't care about.
  32.   metric_relabel_configs:
  33.     [ - <relabel_config> ... ]
  35.   # How frequent to truncate the WAL for this integration.
  36.   [wal_truncate_frequency: <duration> | default = "60m"]
  38.   # Monitor the exporter itself and include those metrics in the results.
  39.   [include_exporter_metrics: <boolean> | default = false]
  41.   # Optionally defines the the list of enabled-by-default collectors.
  42.   # Anything not provided in the list below will be disabled by default,
  43.   # but requires at least one element to be treated as defined.
  44.   #
  45.   # This is useful if you have a very explicit set of collectors you wish
  46.   # to run.
  47.   set_collectors:
  48.     - [<string>]
  50.   # Additional collectors to enable on top of the default set of enabled
  51.   # collectors or on top of the list provided by set_collectors.
  52.   #
  53.   # This is useful if you have a few collectors you wish to run that are
  54.   # not enabled by default, but do not want to explicitly provide an entire
  55.   # list through set_collectors.
  56.   enable_collectors:
  57.     - [<string>]
  59.   # Additional collectors to disable on top of the default set of disabled
  60.   # collectors. Takes precedence over enable_collectors.
  61.   #
  62.   # This is useful if you have a few collectors you do not want to run that
  63.   # are enabled by default, but do not want to explicitly provide an entire
  64.   # list through set_collectors.
  65.   disable_collectors:
  66.     - [<string>]
  68.   # procfs mountpoint.
  69.   [procfs_path: <string> | default = "/proc"]
  71.   # sysfs mountpoint.
  72.   [sysfs_path: <string> | default = "/sys"]
  74.   # rootfs mountpoint. If running in docker, the root filesystem of the host
  75.   # machine should be mounted and this value should be changed to the mount
  76.   # directory.
  77.   [rootfs_path: <string> | default = "/"]
  79.   # Expose expensive bcache priority stats.
  80.   [enable_bcache_priority_stats: <boolean>]
  82.   # Regexp of `bugs` field in cpu info to filter.
  83.   [cpu_bugs_include: <string>]
  85.   # Enable the node_cpu_guest_seconds_total metric.
  86.   [enable_cpu_guest_seconds_metric: <boolean> | default = true]
  88.   # Enable the cpu_info metric for the cpu collector.
  89.   [enable_cpu_info_metric: <boolean> | default = true]
  91.   # Regexp of `flags` field in cpu info to filter.
  92.   [cpu_flags_include: <string>]
  94.   # Regexmp of devices to ignore for diskstats.
  95.   [diskstats_ignored_devices: <string> | default = "^(ram|loop|fd|(h|s|v|xv)d[a-z]|nvme\\d+n\\d+p)\\d+$"]
  97.   # Regexp of ethtool devices to exclude (mutually exclusive with ethtool_device_include)
  98.   [ethtool_device_exclude: <string>]
  100.   # Regexp of ethtool devices to include (mutually exclusive with ethtool_device_exclude)
  101.   [ethtool_device_include: <string>]
  103.   # Regexp of ethtool stats to include.
  104.   [ethtool_metrics_include: <string> | default = ".*"]
  106.   # Regexp of mount points to ignore for filesystem collector.
  107.   [filesystem_mount_points_exclude: <string> | default = "^/(dev|proc|sys|var/lib/docker/.+)($|/)"]
  109.   # Regexp of filesystem types to ignore for filesystem collector.
  110.   [filesystem_fs_types_exclude: <string> | default = "^(autofs|binfmt_misc|bpf|cgroup2?|configfs|debugfs|devpts|devtmpfs|fusectl|hugetlbfs|iso9660|mqueue|nsfs|overlay|proc|procfs|pstore|rpc_pipefs|securityfs|selinuxfs|squashfs|sysfs|tracefs)$"]
  112.   # How long to wait for a mount to respond before marking it as stale.
  113.   [filesystem_mount_timeout: <duration> | default = "5s"]
  115.   # Array of IPVS backend stats labels.
  116.   #
  117.   # The default is [local_address, local_port, remote_address, remote_port, proto, local_mark].
  118.   ipvs_backend_labels:
  119.     [- <string>]
  121.   # NTP server to use for ntp collector
  122.   [ntp_server: <string> | default = "127.0.0.1"]
  124.   # NTP protocol version
  125.   [ntp_protocol_version: <int> | default = 4]
  127.   # Certify that the server address is not a public ntp server.
  128.   [ntp_server_is_local: <boolean> | default = false]
  130.   # IP TTL to use wile sending NTP query.
  131.   [ntp_ip_ttl: <int> | default = 1]
  133.   # Max accumulated distance to the root.
  134.   [ntp_max_distance: <duration> | default = "3466080us"]
  136.   # Offset between local clock and local ntpd time to tolerate.
  137.   [ntp_local_offset_tolerance: <duration> | default = "1ms"]
  139.   # Regexp of net devices to ignore for netclass collector.
  140.   [netclass_ignored_devices: <string> | default = "^$"]
  142.   # Ignore net devices with invalid speed values. This will default to true in
  143.   # node_exporter 2.0.
  144.   [netclass_ignore_invalid_speed_device: <boolean> | default = false]
  146.   # Enable collecting address-info for every device.
  147.   [netdev_address_info: <boolean>]
  149.   # Regexp of net devices to exclude (mutually exclusive with include)
  150.   [netdev_device_exclude: <string> | default = ""]
  152.   # Regexp of net devices to include (mutually exclusive with exclude)
  153.   [netdev_device_include: <string> | default = ""]
  155.   # Regexp of fields to return for netstat collector.
  156.   [netstat_fields: <string> | default = "^(.*_(InErrors|InErrs)|Ip_Forwarding|Ip(6|Ext)_(InOctets|OutOctets)|Icmp6?_(InMsgs|OutMsgs)|TcpExt_(Listen.*|Syncookies.*|TCPSynRetrans|TCPTimeouts)|Tcp_(ActiveOpens|InSegs|OutSegs|OutRsts|PassiveOpens|RetransSegs|CurrEstab)|Udp6?_(InDatagrams|OutDatagrams|NoPorts|RcvbufErrors|SndbufErrors))$"]
  158.   # List of CPUs from which perf metrics should be collected.
  159.   [perf_cpus: <string> | default = ""]
  161.   # Array of perf tracepoints that should be collected.
  162.   perf_tracepoint:
  163.     [- <string>]
  165.   # Regexp of power supplies to ignore for the powersupplyclass collector.
  166.   [powersupply_ignored_supplies: <string> | default = "^$"]
  168.   # Path to runit service directory.
  169.   [runit_service_dir: <string> | default = "/etc/service"]
  171.   # XML RPC endpoint for the supervisord collector.
  172.   #
  173.   # Setting SUPERVISORD_URL in the environment will override the default value.
  174.   # An explicit value in the YAML config takes precedence over the environment
  175.   # variable.
  176.   [supervisord_url: <string> | default = "http://localhost:9001/RPC2"]
  178.   # Regexp of systemd units to include. Units must both match include and not
  179.   # match exclude to be collected.
  180.   [systemd_unit_include: <string> | default = ".+"]
  182.   # Regexp of systemd units to exclude. Units must both match include and not
  183.   # match exclude to be collected.
  184.   [systemd_unit_exclude: <string> | default = ".+\\.(automount|device|mount|scope|slice)"]
  186.   # Enables service unit tasks metrics unit_tasks_current and unit_tasks_max
  187.   [systemd_enable_task_metrics: <boolean> | default = false]
  189.   # Enables service unit metric service_restart_total
  190.   [systemd_enable_restarts_metrics: <boolean> | default = false]
  192.   # Enables service unit metric unit_start_time_seconds
  193.   [systemd_enable_start_time_metrics: <boolean> | default = false]
  195.   # Regexp of tapestats devices to ignore.
  196.   [tapestats_ignored_devices: <string> | default = "^$"]
  198.   # Directory to read *.prom files from for the textfile collector.
  199.   [textfile_directory: <string> | default = ""]
  201.   # Regexp of fields to return for the vmstat collector.
  202.   [vmstat_fields: <string> | default = "^(oom_kill|pgpg|pswp|pg.*fault).*"]

node_exporter 自定义 collectors

您可以在 integrations node_export 配置中，通过设置和修改 set_collectors enable_collectors disable_collectors，以控制哪些 collector 生效。

  1. const (
  2.   CollectorARP          = "arp"
  3.     CollectorBCache       = "bcache"
  4.     CollectorBTRFS        = "btrfs"
  5.     CollectorBonding      = "bonding"
  6.     CollectorBootTime     = "boottime"
  7.     CollectorBuddyInfo    = "buddyinfo"
  8.     CollectorCPU          = "cpu"
  9.     CollectorCPUFreq      = "cpufreq"
  10.     CollectorConntrack    = "conntrack"
  11.     CollectorDMI          = "dmi"
  12.     CollectorDRBD         = "drbd"
  13.     CollectorDRM          = "drm"
  14.     CollectorDevstat      = "devstat"
  15.     CollectorDiskstats    = "diskstats"
  16.     CollectorEDAC         = "edac"
  17.     CollectorEntropy      = "entropy"
  18.     CollectorEthtool      = "ethtool"
  19.     CollectorExec         = "exec"
  20.     CollectorFibrechannel = "fibrechannel"
  21.     CollectorFileFD       = "filefd"
  22.     CollectorFilesystem   = "filesystem"
  23.     CollectorHWMon        = "hwmon"
  24.     CollectorIPVS         = "ipvs"
  25.     CollectorInfiniband   = "infiniband"
  26.     CollectorInterrupts   = "interrupts"
  27.     CollectorKSMD         = "ksmd"
  28.     CollectorLnstat       = "lnstat"
  29.     CollectorLoadAvg      = "loadavg"
  30.     CollectorLogind       = "logind"
  31.     CollectorMDADM        = "mdadm"
  32.     CollectorMeminfo      = "meminfo"
  33.     CollectorMeminfoNuma  = "meminfo_numa"
  34.     CollectorMountstats   = "mountstats"
  35.     CollectorNFS          = "nfs"
  36.     CollectorNFSD         = "nfsd"
  37.     CollectorNTP          = "ntp"
  38.     CollectorNVME         = "nvme"
  39.     CollectorNetclass     = "netclass"
  40.     CollectorNetdev       = "netdev"
  41.     CollectorNetstat      = "netstat"
  42.     CollectorNetworkRoute = "network_route"
  43.     CollectorOS           = "os"
  44.     CollectorPerf         = "perf"
  45.     CollectorPowersuppply = "powersupplyclass"
  46.     CollectorPressure     = "pressure"
  47.     CollectorProcesses    = "processes"
  48.     CollectorQDisc        = "qdisc"
  49.     CollectorRAPL         = "rapl"
  50.     CollectorRunit        = "runit"
  51.     CollectorSchedstat    = "schedstat"
  52.     CollectorSockstat     = "sockstat"
  53.     CollectorSoftnet      = "softnet"
  54.     CollectorStat         = "stat"
  55.     CollectorSupervisord  = "supervisord"
  56.     CollectorSystemd      = "systemd"
  57.     CollectorTCPStat      = "tcpstat"
  58.     CollectorTapestats    = "tapestats"
  59.     CollectorTextfile     = "textfile"
  60.     CollectorThermal      = "thermal"
  61.     CollectorThermalzone  = "thermal_zone"
  62.     CollectorTime         = "time"
  63.     CollectorTimex        = "timex"
  64.     CollectorUDPQueues    = "udp_queues"
  65.     CollectorUname        = "uname"
  66.     CollectorVMStat       = "vmstat"
  67.     CollectorWiFi         = "wifi"
  68.     CollectorXFS          = "xfs"
  69.     CollectorZFS          = "zfs"
  70.     CollectorZoneinfo     = "zoneinfo"
  71. )