Exporter简单入门 - Alfons's Blog

各个采集端(Exporter)是 Prometheus 的数据来源，在看了多个官方的 exporter 后，总结了一下。

如何编写 Exporter

编写 Exporter，有许多种语言的实现，下面以 Go 框架为基础。

总体概括，Exporter 编写主要有下面三个步骤：

「1」创建采集项 Collector。Collector 核心为 Collect 方法，指标项结果处理后，通过 ch 管道传出

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
type Collector struct {}

func (c *Collector) Describe(ch chan<- *prometheus.Desc) {}

func (c *Collector) Collect(ch chan<- prometheus.Metric) {
    desc := prometheus.NewDesc(
        prometheus.BuildFQName("node", "cpu", "info"),
        "Show CPU information.",
        []string{"label_1", "label_2", "label_3"},
        prometheus.Labels{"Hello": "World"},
    )
    ch <- prometheus.MustNewConstMetric(
        desc,
        prometheus.CounterValue,
        1.0,
        "1.0", "2.0", "3.0",
    )
}

「2」注册采集项对象。通过框架标准方法注册采集项 Collector

1
2
3
// Create a non-global registry.
reg := prometheus.NewRegistry()
_ = reg.Register(c)

「3」绑定HttpHandler，并启动web服务

1
2
3
4
// Expose metrics and custom registry via an HTTP server
// using the HandleFor function. "/metrics" is the usual endpoint for that.
http.Handle("/node_metrics", promhttp.HandlerFor(reg, promhttp.HandlerOpts{Registry: reg}))
log.Fatal(http.ListenAndServe(":8080", nil))  

Exporter 本质上只是一个web服务。Prometheus 会通过配置中的 API，对 Exporter 发起请求，Exporter 对应的借口进行响应。最终通过触发 Collector.Collect 函数进行采集。

在 Collector.Collect 函数中，传入的参数是：ch chan<- prometheus.Metric。采集指标通过这个管道往调用层传输，最终输出到API结果中。

采集指标 prometheus.Metric 也是有一定的结构、类型的。

采集指标主要使用到的类型包括：Gauge、Counter。另外两个 Histogram、Summary，一般不会使用到，这里不讲。

采集指标主要由 描述信息、值、label、固定label 组成

初始化指标实例的方法有很多，目前比较主流的方式是：先初始化描述信息、在执行 Collector.Collect 动作时，填充其他信息。

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
desc := prometheus.NewDesc(
    prometheus.BuildFQName("node", "cpu", "info"),
    "Show CPU information.",
    []string{"label_1", "label_2", "label_3"},
    prometheus.Labels{"Hello": "World"},
)

ch <- prometheus.MustNewConstMetric(
    desc,
    prometheus.CounterValue,
    1.0,
    "1.0", "2.0", "3.0",
)

prometheus.NewDesc 具体参数如下：

1
func NewDesc(fqName, help string, variableLabels []string, constLabels Labels) *Desc

fqName - 采集指标名称，通过 prometheus.BuildFQName 函数生成，也可以自定义
help - 采集指标的 help 信息
variableLabels - 采集指标的 label
constLabels - 固定的采集指标 label

prometheus.MustNewConstMetric 具体参数如下：

1
func MustNewConstMetric(desc *Desc, valueType ValueType, value float64, labelValues ...string) Metric

desc - 描述信息
valueType - 指标值的类型
value - 指标值
labelValues - label 的值

如上面的例子，将输出：

1
2
3
# HELP node_cpu_info Show CPU information.
# TYPE node_cpu_info counter
node_cpu_info{Hello="World",label_1="1.0",label_2="2.0",label_3="3.0"} 1

完整事例代码如下：

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
package main

import (
    "github.com/prometheus/client_golang/prometheus"
    "github.com/prometheus/client_golang/prometheus/promhttp"
    "log"
    "net/http"
    "runtime"
)

type Metrics struct {
    cpuTemp        prometheus.Gauge
    goRuntime      prometheus.GaugeFunc
    hdFailures     prometheus.Counter
    hdFailuresFunc prometheus.CounterFunc

    //metrics prometheus.Metric
}

func (c *Metrics) Describe(ch chan<- *prometheus.Desc) {
    ch <- c.cpuTemp.Desc()
    ch <- c.hdFailures.Desc()
    ch <- c.goRuntime.Desc()
}

func (c *Metrics) Collect(ch chan<- prometheus.Metric) {
    c.cpuTemp.Set(65.3)
    ch <- c.cpuTemp

    c.goRuntime.Collect(ch)

    c.hdFailures.Inc()
    ch <- c.hdFailures

    c.hdFailuresFunc.Collect(ch)

    ch <- prometheus.MustNewConstMetric(
        prometheus.NewDesc(
            prometheus.BuildFQName("node", "cpu", "info"),
            "Show CPU information.",
            []string{"label_1", "label_2", "label_3"},
            prometheus.Labels{"Hello": "World"},
        ),
        prometheus.CounterValue,
        1.0,
        "1.0", "2.0", "3.0",
    )
}

func main() {
    m := &Metrics{
        cpuTemp: prometheus.NewGauge(
            prometheus.GaugeOpts{
                Name: "cpu_temperature_celsius",
                Help: "Current temperature of the CPU.",
            },
        ),
        goRuntime: prometheus.NewGaugeFunc(
            prometheus.GaugeOpts{
                Subsystem: "runtime",
                Name:      "goroutines_count",
                Help:      "Number of goroutines that currently exist.",
            },
            func() float64 { return float64(runtime.NumGoroutine()) },
        ),

        hdFailures: prometheus.NewCounter(
            prometheus.CounterOpts{
                Name: "hd_errors_total",
                Help: "Number of hard-disk errors.",
            },
        ),
        hdFailuresFunc: prometheus.NewCounterFunc(
            prometheus.CounterOpts{
                Name: "hd_errors_total_func",
                Help: "Number of hard-disk errors.",
            },
            func() float64 { return 1.0 },
        ),
    }

    // Create a non-global registry.
    reg := prometheus.NewRegistry()
    _ = reg.Register(m)

    //// Create new metrics and register them using the custom registry.
    //m := NewMetrics(reg)

    //// Set values for the new created metrics.
    //m.cpuTemp.Set(65.3)
    //m.hdFailures.With(prometheus.Labels{"device": "/dev/sda"}).Inc()

    // Expose metrics and custom registry via an HTTP server
    // using the HandleFor function. "/metrics" is the usual endpoint for that.
    http.Handle("/metrics", promhttp.HandlerFor(reg, promhttp.HandlerOpts{Registry: reg}))
    log.Fatal(http.ListenAndServe(":8080", nil))
}

总结

Prometheus 的 Exporter 就是一个 Web 服务端，Prometheus 通过定时访问对应的 API 进行指标项的采集
Exporter 的编写按照上述的 “三板斧” 进行
最终采集指标的值，通过管道传输到调用层

参考

client_golang

Contents

如何编写 Exporter

总结

参考