CPU
CPU 采集器用于系统 CPU 使用率等指标的采集。
配置¶
采集器配置¶
成功安装 DataKit 并启动后,会默认开启 CPU 采集器,无需手动开启。
进入 DataKit 安装目录下的 conf.d/samples 目录,复制 cpu.conf.sample 并命名为 cpu.conf。示例如下:
[[inputs.cpu]]
## Collect interval, default is 10 seconds. (optional)
interval = '10s'
## Collect CPU usage per core, default is false. (optional)
percpu = false
## Setting disable_temperature_collect to false will collect cpu temperature stats for linux. (deprecated)
# disable_temperature_collect = false
## Enable to collect core temperature data.
enable_temperature = true
## Enable gets average load information every five seconds.
enable_load5s = true
[inputs.cpu.tags]
# some_tag = "some_value"
# more_tag = "some_other_value"
配置好后,重启 DataKit 即可。
可通过 ConfigMap 方式注入采集器配置 或 配置 ENV_DATAKIT_INPUTS 开启采集器。
也支持以环境变量的方式修改配置参数(需要在 ENV_DEFAULT_ENABLED_INPUTS 中加为默认采集器):
-
ENV_INPUT_CPU_INTERVAL
采集器重复间隔时长
字段类型: Duration
采集器配置字段:
interval默认值:
10s -
ENV_INPUT_CPU_PERCPU
采集每一个 CPU 核
字段类型: Boolean
采集器配置字段:
percpu默认值: false
-
ENV_INPUT_CPU_ENABLE_TEMPERATURE
采集 CPU 温度
字段类型: Boolean
采集器配置字段:
enable_temperature默认值:
true -
ENV_INPUT_CPU_ENABLE_LOAD5S
每五秒钟获取一次平均负载信息
字段类型: Boolean
采集器配置字段:
enable_load5s默认值:
false -
ENV_INPUT_CPU_TAGS
自定义标签。如果配置文件有同名标签,将会覆盖它
字段类型: String
采集器配置字段:
tags示例:
tag1=value1,tag2=value2
指标¶
以下所有数据采集,默认会追加名为 host 的全局 tag(tag 值为 DataKit 所在主机名),也可以在配置中通过 [inputs.cpu.tags] 指定其它标签:
cpu¶
CPU usage, load, and optional core temperature metrics derived from host CPU time deltas and sensor readings.
| Tags & Fields | Description |
|---|---|
| cpu ( tag) |
CPU identifier. cpu-total is the aggregate across all CPUs; per-core values are emitted only when percpu is enabled. |
| host ( tag) |
System hostname. |
| core_temperature | Average CPU core temperature across detected core temperature sensors, in degrees Celsius. Linux only. Type: float | (gauge) Unit: temperature,C |
| load5s | Five-second CPU load estimate as a dimensionless host load value. Type: int | (gauge) Unit: N/A |
| usage_guest | Percentage of CPU time spent running virtual CPUs for guest operating systems during the collection interval. Linux only. Type: float | (gauge) Unit: percent,percent |
| usage_guest_nice | Percentage of CPU time spent running low-priority virtual CPUs for guest operating systems during the collection interval. Linux only, kernel 3.2.0 or later. Type: float | (gauge) Unit: percent,percent |
| usage_idle | Percentage of CPU time spent idle during the collection interval. Type: float | (gauge) Unit: percent,percent |
| usage_iowait | Percentage of CPU time spent waiting for I/O completion during the collection interval. Linux only. Type: float | (gauge) Unit: percent,percent |
| usage_irq | Percentage of CPU time spent servicing hardware interrupts during the collection interval. Type: float | (gauge) Unit: percent,percent |
| usage_nice | Percentage of CPU time spent running low-priority nice user processes during the collection interval. Linux only. Type: float | (gauge) Unit: percent,percent |
| usage_softirq | Percentage of CPU time spent servicing software interrupts during the collection interval. Linux only. Type: float | (gauge) Unit: percent,percent |
| usage_steal | Percentage of CPU time stolen by other operating systems when running in a virtualized environment. Linux only. Type: float | (gauge) Unit: percent,percent |
| usage_system | Percentage of CPU time spent in system mode during the collection interval. Type: float | (gauge) Unit: percent,percent |
| usage_total | Total active CPU usage percentage during the collection interval, computed from user, system, nice, iowait, irq, softirq, steal, guest, and guest_nice time. Type: float | (gauge) Unit: percent,percent |
| usage_user | Percentage of CPU time spent in user mode during the collection interval. Type: float | (gauge) Unit: percent,percent |