{ "componentChunkName": "component---src-templates-issues-tsx", "path": "/issues/400", "result": {"data":{"issuesJson":{"id":"36d317e3-d4bb-52df-b30d-e29e1e2fed11","title":"prometheus/node_exporter: 用于机器指标的 Exporter","number":400,"bodyHTML":"
\n



\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Reposprometheus/node_exporter
Views\"views\"
Stars\"stars\"
Forks\"forks\"
License\"license\"
UpdatedAt\"last-commit\"
CreatedAt\"create-at\"
\n\n

\"\"

\n\n
\n

Node exporter

\n

\"CircleCI\"
\n\"bsd
\n\"golangci-lint
\n\"Docker
\n\"Docker
\n\"Go

\n

Prometheus exporter for hardware and OS metrics exposed by *NIX kernels, written
\nin Go with pluggable metric collectors.

\n

The Windows exporter is recommended for Windows users.
\nTo expose NVIDIA GPU metrics, prometheus-dcgm
\n
\ncan be used.

\n

Installation and Usage

\n

If you are new to Prometheus and node_exporter there is a simple step-by-step guide.

\n

The node_exporter listens on HTTP port 9100 by default. See the --help output for more options.

\n

Ansible

\n

For automated installs with Ansible, there is the Prometheus Community role.

\n

Docker

\n

The node_exporter is designed to monitor the host system. Deploying in containers requires
\nextra care in order to avoid monitoring the container itself.

\n

For situations where containerized deployment is needed, some extra flags must be used to allow
\nthe node_exporter access to the host namespaces.

\n

Be aware that any non-root mount points you want to monitor will need to be bind-mounted
\ninto the container.

\n

If you start container for host monitoring, specify path.rootfs argument.
\nThis argument must match path in bind-mount of host root. The node_exporter will use
\npath.rootfs as prefix to access host filesystem.

\n
docker run -d \\\n  --net=\"host\" \\\n  --pid=\"host\" \\\n  -v \"/:/host:ro,rslave\" \\\n  quay.io/prometheus/node-exporter:latest \\\n  --path.rootfs=/host
\n

For Docker compose, similar flag changes are needed.

\n
---\nversion: '3.8'\n\nservices:\n  node_exporter:\n    image: quay.io/prometheus/node-exporter:latest\n    container_name: node_exporter\n    command:\n      - '--path.rootfs=/host'\n    network_mode: host\n    pid: host\n    restart: unless-stopped\n    volumes:\n      - '/:/host:ro,rslave'
\n

On some systems, the timex collector requires an additional Docker flag,
\n--cap-add=SYS_TIME, in order to access the required syscalls.

\n

Collectors

\n

There is varying support for collectors on each operating system. The tables
\nbelow list all existing collectors and the supported systems.

\n

Collectors are enabled by providing a --collector.<name> flag.
\nCollectors that are enabled by default can be disabled by providing a --no-collector.<name> flag.
\nTo enable only some specific collector(s), use --collector.disable-defaults --collector.<name> ....

\n

Include & Exclude flags

\n

A few collectors can be configured to include or exclude certain patterns using dedicated flags. The exclude flags are used to indicate \"all except\", while the include flags are used to say \"none except\". Note that these flags are mutually exclusive on collectors that support both.

\n

Example:

\n
--collector.filesystem.mount-points-exclude=^/(dev|proc|sys|var/lib/docker/.+|var/lib/kubelet/.+)($|/)
\n

List:

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
CollectorScopeInclude FlagExclude Flag
arpdevice--collector.arp.device-include--collector.arp.device-exclude
cpubugs--collector.cpu.info.bugs-includeN/A
cpuflags--collector.cpu.info.flags-includeN/A
diskstatsdevice--collector.diskstats.device-include--collector.diskstats.device-exclude
ethtooldevice--collector.ethtool.device-include--collector.ethtool.device-exclude
ethtoolmetrics--collector.ethtool.metrics-includeN/A
filesystemfs-types--collector.filesystem.fs-types-include--collector.filesystem.fs-types-exclude
filesystemmount-points--collector.filesystem.mount-points-include--collector.filesystem.mount-points-exclude
hwmonchip--collector.hwmon.chip-include--collector.hwmon.chip-exclude
hwmonsensor--collector.hwmon.sensor-include--collector.hwmon.sensor-exclude
interruptsname--collector.interrupts.name-include--collector.interrupts.name-exclude
netdevdevice--collector.netdev.device-include--collector.netdev.device-exclude
qdiskdevice--collector.qdisk.device-include--collector.qdisk.device-exclude
slabinfoslab-names--collector.slabinfo.slabs-include--collector.slabinfo.slabs-exclude
sysctlall--collector.sysctl.includeN/A
systemdunit--collector.systemd.unit-include--collector.systemd.unit-exclude
\n

Enabled by default

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
NameDescriptionOS
arpExposes ARP statistics from /proc/net/arp.Linux
bcacheExposes bcache statistics from /sys/fs/bcache/.Linux
bondingExposes the number of configured and active slaves of Linux bonding interfaces.Linux
btrfsExposes btrfs statisticsLinux
boottimeExposes system boot time derived from the kern.boottime sysctl.Darwin, Dragonfly, FreeBSD, NetBSD, OpenBSD, Solaris
conntrackShows conntrack statistics (does nothing if no /proc/sys/net/netfilter/ present).Linux
cpuExposes CPU statisticsDarwin, Dragonfly, FreeBSD, Linux, Solaris, OpenBSD
cpufreqExposes CPU frequency statisticsLinux, Solaris
diskstatsExposes disk I/O statistics.Darwin, Linux, OpenBSD
dmiExpose Desktop Management Interface (DMI) info from /sys/class/dmi/id/Linux
edacExposes error detection and correction statistics.Linux
entropyExposes available entropy.Linux
execExposes execution statistics.Dragonfly, FreeBSD
fibrechannelExposes fibre channel information and statistics from /sys/class/fc_host/.Linux
filefdExposes file descriptor statistics from /proc/sys/fs/file-nr.Linux
filesystemExposes filesystem statistics, such as disk space used.Darwin, Dragonfly, FreeBSD, Linux, OpenBSD
hwmonExpose hardware monitoring and sensor data from /sys/class/hwmon/.Linux
infinibandExposes network statistics specific to InfiniBand and Intel OmniPath configurations.Linux
ipvsExposes IPVS status from /proc/net/ip_vs and stats from /proc/net/ip_vs_stats.Linux
loadavgExposes load average.Darwin, Dragonfly, FreeBSD, Linux, NetBSD, OpenBSD, Solaris
mdadmExposes statistics about devices in /proc/mdstat (does nothing if no /proc/mdstat present).Linux
meminfoExposes memory statistics.Darwin, Dragonfly, FreeBSD, Linux, OpenBSD
netclassExposes network interface info from /sys/class/net/Linux
netdevExposes network interface statistics such as bytes transferred.Darwin, Dragonfly, FreeBSD, Linux, OpenBSD
netisrExposes netisr statisticsFreeBSD
netstatExposes network statistics from /proc/net/netstat. This is the same information as netstat -s.Linux
nfsExposes NFS client statistics from /proc/net/rpc/nfs. This is the same information as nfsstat -c.Linux
nfsdExposes NFS kernel server statistics from /proc/net/rpc/nfsd. This is the same information as nfsstat -s.Linux
nvmeExposes NVMe info from /sys/class/nvme/Linux
osExpose OS release info from /etc/os-release or /usr/lib/os-releaseany
powersupplyclassExposes Power Supply statistics from /sys/class/power_supplyLinux
pressureExposes pressure stall statistics from /proc/pressure/.Linux (kernel 4.20+ and/or CONFIG_PSI)
raplExposes various statistics from /sys/class/powercap.Linux
schedstatExposes task scheduler statistics from /proc/schedstat.Linux
selinuxExposes SELinux statistics.Linux
sockstatExposes various statistics from /proc/net/sockstat.Linux
softnetExposes statistics from /proc/net/softnet_stat.Linux
statExposes various statistics from /proc/stat. This includes boot time, forks and interrupts.Linux
tapestatsExposes statistics from /sys/class/scsi_tape.Linux
textfileExposes statistics read from local disk. The --collector.textfile.directory flag must be set.any
thermalExposes thermal statistics like pmset -g therm.Darwin
thermal_zoneExposes thermal zone & cooling device statistics from /sys/class/thermal.Linux
timeExposes the current system time.any
timexExposes selected adjtimex(2) system call stats.Linux
udp_queuesExposes UDP total lengths of the rx_queue and tx_queue from /proc/net/udp and /proc/net/udp6.Linux
unameExposes system information as provided by the uname system call.Darwin, FreeBSD, Linux, OpenBSD
vmstatExposes statistics from /proc/vmstat.Linux
watchdogExposes statistics from /sys/class/watchdogLinux
xfsExposes XFS runtime statistics.Linux (kernel 4.4+)
zfsExposes ZFS performance statistics.FreeBSD, Linux, Solaris
\n

Disabled by default

\n

node_exporter also implements a number of collectors that are disabled by default. Reasons for this vary by
\ncollector, and may include:

\n\n

You can enable additional collectors as desired by adding them to your
\ninit system's or service supervisor's startup configuration for
\nnode_exporter but caution is advised. Enable at most one at a time,
\ntesting first on a non-production system, then by hand on a single
\nproduction node. When enabling additional collectors, you should
\ncarefully monitor the change by observing the scrape_duration_seconds metric to ensure that collection completes
\nand does not time out. In addition, monitor the
\nscrape_samples_post_metric_relabeling metric to see the changes in
\ncardinality.

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
NameDescriptionOS
buddyinfoExposes statistics of memory fragments as reported by /proc/buddyinfo.Linux
cgroupsA summary of the number of active and enabled cgroupsLinux
cpu_vulnerabilitiesExposes CPU vulnerability information from sysfs.Linux
devstatExposes device statisticsDragonfly, FreeBSD
drmExpose GPU metrics using sysfs / DRM, amdgpu is the only driver which exposes this information through DRMLinux
drbdExposes Distributed Replicated Block Device statistics (to version 8.4)Linux
ethtoolExposes network interface information and network driver statistics equivalent to ethtool, ethtool -S, and ethtool -i.Linux
interruptsExposes detailed interrupts statistics.Linux, OpenBSD
ksmdExposes kernel and system statistics from /sys/kernel/mm/ksm.Linux
lnstatExposes stats from /proc/net/stat/.Linux
logindExposes session counts from logind.Linux
meminfo_numaExposes memory statistics from /sys/devices/system/node/node[0-9]*/meminfo, /sys/devices/system/node/node[0-9]*/numastat.Linux
mountstatsExposes filesystem statistics from /proc/self/mountstats. Exposes detailed NFS client statistics.Linux
network_routeExposes the routing table as metricsLinux
perfExposes perf based metrics (Warning: Metrics are dependent on kernel configuration and settings).Linux
processesExposes aggregate process statistics from /proc.Linux
qdiscExposes queuing discipline statisticsLinux
slabinfoExposes slab statistics from /proc/slabinfo. Note that permission of /proc/slabinfo is usually 0400, so set it appropriately.Linux
softirqsExposes detailed softirq statistics from /proc/softirqs.Linux
sysctlExpose sysctl values from /proc/sys. Use --collector.sysctl.include(-info) to configure.Linux
systemdExposes service and system status from systemd.Linux
tcpstatExposes TCP connection status information from /proc/net/tcp and /proc/net/tcp6. (Warning: the current version has potential performance issues in high load situations.)Linux
wifiExposes WiFi device and station statistics.Linux
xfrmExposes statistics from /proc/net/xfrm_statLinux
zoneinfoExposes NUMA memory zone metrics.Linux
\n

Deprecated

\n

These collectors are deprecated and will be removed in the next major release.

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
NameDescriptionOS
ntpExposes local NTP daemon health to check timeany
runitExposes service status from runit.any
supervisordExposes service status from supervisord.any
\n

Perf Collector

\n

The perf collector may not work out of the box on some Linux systems due to kernel
\nconfiguration and security settings. To allow access, set the following sysctl
\nparameter:

\n
sysctl -w kernel.perf_event_paranoid=X\n
\n\n

Depending on the configured value different metrics will be available, for most
\ncases 0 will provide the most complete set. For more information see man 2 perf_event_open.

\n

By default, the perf collector will only collect metrics of the CPUs that
\nnode_exporter is running on (ie
\nruntime.NumCPU. If this is
\ninsufficient (e.g. if you run node_exporter with its CPU affinity set to
\nspecific CPUs), you can specify a list of alternate CPUs by using the
\n--collector.perf.cpus flag. For example, to collect metrics on CPUs 2-6, you
\nwould specify: --collector.perf --collector.perf.cpus=2-6. The CPU
\nconfiguration is zero indexed and can also take a stride value; e.g.
\n--collector.perf --collector.perf.cpus=1-10:5 would collect on CPUs
\n1, 5, and 10.

\n

The perf collector is also able to collect
\ntracepoint
\ncounts when using the --collector.perf.tracepoint flag. Tracepoints can be
\nfound using perf list or
\nfrom debugfs. And example usage of this would be
\n--collector.perf.tracepoint=\"sched:sched_process_exec\".

\n

Sysctl Collector

\n

The sysctl collector can be enabled with --collector.sysctl. It supports exposing numeric sysctl values
\nas metrics using the --collector.sysctl.include flag and string values as info metrics by using the
\n--collector.sysctl.include-info flag. The flags can be repeated. For sysctl with multiple numeric values,
\nan optional mapping can be given to expose each value as its own metric. Otherwise an index label is used
\nto identify the different fields.

\n

Examples

\n
Numeric values
\n
Single values
\n

Using --collector.sysctl.include=vm.user_reserve_kbytes:
\nvm.user_reserve_kbytes = 131072 -> node_sysctl_vm_user_reserve_kbytes 131072

\n
Multiple values
\n

A sysctl can contain multiple values, for example:

\n
net.ipv4.tcp_rmem = 4096\t131072\t6291456\n
\n

Using --collector.sysctl.include=net.ipv4.tcp_rmem the collector will expose:

\n
node_sysctl_net_ipv4_tcp_rmem{index=\"0\"} 4096\nnode_sysctl_net_ipv4_tcp_rmem{index=\"1\"} 131072\nnode_sysctl_net_ipv4_tcp_rmem{index=\"2\"} 6291456\n
\n

If the indexes have defined meaning like in this case, the values can be mapped to multiple metrics by appending the mapping to the --collector.sysctl.include flag:
\nUsing --collector.sysctl.include=net.ipv4.tcp_rmem:min,default,max the collector will expose:

\n
node_sysctl_net_ipv4_tcp_rmem_min 4096\nnode_sysctl_net_ipv4_tcp_rmem_default 131072\nnode_sysctl_net_ipv4_tcp_rmem_max 6291456\n
\n
String values
\n

String values need to be exposed as info metric. The user selects them by using the --collector.sysctl.include-info flag.

\n
Single values
\n

kernel.core_pattern = core -> node_sysctl_info{key=\"kernel.core_pattern_info\", value=\"core\"} 1

\n
Multiple values
\n

Given the following sysctl:

\n
kernel.seccomp.actions_avail = kill_process kill_thread trap errno trace log allow\n
\n

Setting --collector.sysctl.include-info=kernel.seccomp.actions_avail will yield:

\n
node_sysctl_info{key=\"kernel.seccomp.actions_avail\", index=\"0\", value=\"kill_process\"} 1\nnode_sysctl_info{key=\"kernel.seccomp.actions_avail\", index=\"1\", value=\"kill_thread\"} 1\n...\n
\n

Textfile Collector

\n

The textfile collector is similar to the Pushgateway,
\nin that it allows exporting of statistics from batch jobs. It can also be used
\nto export static metrics, such as what role a machine has. The Pushgateway
\nshould be used for service-level metrics. The textfile module is for metrics
\nthat are tied to a machine.

\n

To use it, set the --collector.textfile.directory flag on the node_exporter commandline. The
\ncollector will parse all files in that directory matching the glob *.prom
\nusing the text
\nformat. Note: Timestamps are not supported.

\n

To atomically push completion time for a cron job:

\n
echo my_batch_job_completion_time $(date +%s) > /path/to/directory/my_batch_job.prom.$$\nmv /path/to/directory/my_batch_job.prom.$$ /path/to/directory/my_batch_job.prom\n
\n

To statically set roles for a machine using labels:

\n
echo 'role{role=\"application_server\"} 1' > /path/to/directory/role.prom.$$\nmv /path/to/directory/role.prom.$$ /path/to/directory/role.prom\n
\n

Filtering enabled collectors

\n

The node_exporter will expose all metrics from enabled collectors by default. This is the recommended way to collect metrics to avoid errors when comparing metrics of different families.

\n

For advanced use the node_exporter can be passed an optional list of collectors to filter metrics. The parameters collect[] and exclude[] can be used multiple times (but cannot be combined). In Prometheus configuration you can use this syntax under the scrape config.

\n

Collect only cpu and meminfo collector metrics:

\n
  params:\n    collect[]:\n      - cpu\n      - meminfo\n
\n

Collect all enabled collector metrics but exclude netdev:

\n
  params:\n    exclude[]:\n      - netdev\n
\n

This can be useful for having different Prometheus servers collect specific metrics from nodes.

\n

Development building and running

\n

Prerequisites:

\n\n

Building:

\n
git clone https://github.com/prometheus/node_exporter.git\ncd node_exporter\nmake build\n./node_exporter <flags>\n
\n

To see all available configuration flags:

\n
./node_exporter -h\n
\n

Running tests

\n
make test\n
\n

TLS endpoint

\n

EXPERIMENTAL

\n

The exporter supports TLS via a new web configuration file.

\n
./node_exporter --web.config.file=web-config.yml
\n

See the exporter-toolkit web-configuration for more details.

","updatedAt":"2025-05-21T16:13:26Z","upvoteCount":null,"author":{"login":"eryajf","avatarUrl":"https://avatars.githubusercontent.com/u/33259379?u=e4a4090a38ac2473aaed4ef9945233636776c6c3&v=4","url":"https://github.com/eryajf"},"category":null,"labels":{"edges":[{"node":{"name":"更多","color":"25B472"}},{"node":{"name":"Prometheus-Exporter","color":"367acb"}},{"node":{"name":"prometheus","color":"3d4dd6"}}]},"comments":{"edges":[]}}},"pageContext":{"number":400,"previous":{"title":"ccfos/nightingale: All-in-One 的云原生监控工具，集合了 Prometheus 和 Grafana 的优点，你可以在 WebUI 上管理和配置告警策略，也可以对分布在多个 Region 的指标、日志、链路追踪数据进行统一的可视化和分析","number":399},"next":{"title":"novuhq/novu: 开源的通知基础设施，可以使用一个 API 管理多个渠道的通知。","number":401}}}, "staticQueryHashes": ["151096407","2861350382"]}