1. 安裝 GO 和依賴包

按照 https://golang.org/doc/install 上的步驟進(jìn)行安裝配置 GO 環(huán)境，創(chuàng)建一個(gè)名為 my_first_exporter 的文件夾。

$?go?mod?init?my_first_exporter?
$?go?get?github.com/prometheus/client_golang?
$?go?get?github.com/joho/godotenv
-->?creates?go.mod?file
-->?Installs?dependency?into?the?go.mod?file

2. 創(chuàng)建入口點(diǎn)和導(dǎo)入依賴包

package?main

import?(
?"github.com/joho/godotenv"
?"github.com/prometheus/client_golang/prometheus"
?"github.com/prometheus/client_golang/prometheus/promhttp"
)

3. 創(chuàng)建 main() 函數(shù)

func?main()

4. 添加 prometheus metrics 端點(diǎn)，并在某個(gè)服務(wù)端口上監(jiān)聽

func?main()?{
???http.Handle("/metrics",?promhttp.Handler())
???log.Fatal(http.ListenAndServe(":9101",?nil))
}

5. 使用 curl 請求外部服務(wù)接口

比如我們這里監(jiān)控的應(yīng)用程序是 MirthConnect，所以我需要進(jìn)行兩個(gè) API 接口調(diào)用：

• 獲取 channel 統(tǒng)計(jì)數(shù)據(jù)
• 獲取 channel id 和名稱映射

curl?-k?--location?--request?GET?'https://apihost/api/channels/statistics'?\
--user?admin:admin

curl?-k?--location?--request?GET?'https://apihost/api/channels/idsAndNames'?\
--user?admin:admin

6. 將 curl 調(diào)用轉(zhuǎn)換為 go http 調(diào)用，并解析結(jié)果

如果你是 Go 新手，這應(yīng)該是最困難的一步。對于我這里的例子，端點(diǎn)返回的是 XML 格式的數(shù)據(jù)，所以我必須用 "encoding/xml" 包來反序列化 XML。轉(zhuǎn)換成功后意味著我的 GO 程序可以執(zhí)行和 curl 命令一樣的 API 調(diào)用。

7. 聲明 Prometheus metrics

在 Prometheus 中，每個(gè) metric 指標(biāo)都由以下幾個(gè)部分組成：metric name/metric label values/metric help text/metric type/measurement ，例如：

Example:
#?HELP?promhttp_metric_handler_requests_total?Total?number?of?scrapes?by?HTTP?status?code.
#?TYPE?promhttp_metric_handler_requests_total?counter
promhttp_metric_handler_requests_total{code=”200"}?1.829159e+06
promhttp_metric_handler_requests_total{code=”500"}?0
promhttp_metric_handler_requests_total{code=”503"}?0

對于應(yīng)用 scrapers，我們將定義 Prometheus metrics 描述信息，其中包括 metric 名稱、metric label 標(biāo)簽以及 metric 幫助信息。

messagesReceived?=?prometheus.NewDesc(
?prometheus.BuildFQName(namespace,?"",?"messages_received_total"),
?"How?many?messages?have?been?received?(per?channel).",
?[]string{"channel"},?nil,
)

8. 定義一個(gè)結(jié)構(gòu)體實(shí)現(xiàn) Prometheus 的 Collector 接口

Prometheus 的 client 庫提供了實(shí)現(xiàn)自定義 Exportor 的接口，Collector 接口定義了兩個(gè)方法 Describe 和 Collect，實(shí)現(xiàn)這兩個(gè)方法就可以暴露自定義的數(shù)據(jù)：

• Describe(chan<- *Desc)
• Collect(chan<- Metric)

如下所示：

type?Exporter?struct?{
?mirthEndpoint,?mirthUsername,?mirthPassword?string
}

func?NewExporter(mirthEndpoint?string,?mirthUsername?string,?mirthPassword?string)?*Exporter?{
?return?&Exporter{
??mirthEndpoint:?mirthEndpoint,
??mirthUsername:?mirthUsername,
??mirthPassword:?mirthPassword,
?}
}
func?(e?*Exporter)?Describe(ch?chan<-?*prometheus.Desc)?{
}
func?(e?*Exporter)?Collect(ch?chan<-?prometheus.Metric)?{
}

9. 在 Describe 函數(shù)中，把第7步的 metric 描述信息發(fā)送給它

func?(e?*Exporter)?Describe(ch?chan<-?*prometheus.Desc)?{
?ch?<-?up
?ch?<-?messagesReceived
?ch?<-?messagesFiltered
?ch?<-?messagesQueued
?ch?<-?messagesSent
?ch?<-?messagesErrored
}

10. 將接口調(diào)用邏輯從第6步移到 Collect 函數(shù)中

直接將采集的數(shù)據(jù)發(fā)送到 prometheus.Metric 通道中。

func?(e?*Exporter)?Collect(ch?chan<-?prometheus.Metric)?{
?channelIdNameMap,?err?:=?e.LoadChannelIdNameMap()
?if?err?!=?nil?{
??ch?<-?prometheus.MustNewConstMetric(
???up,?prometheus.GaugeValue,?0,
??)
??log.Println(err)
??return
?}
?ch?<-?prometheus.MustNewConstMetric(
??up,?prometheus.GaugeValue,?1,
?)

?e.HitMirthRestApisAndUpdateMetrics(channelIdNameMap,?ch)
}

當(dāng)執(zhí)行 api 調(diào)用時(shí)，確保使用prometheus.MustNewConstMetric(prometheus.Desc, metric type, measurement)發(fā)送測量值，如果你需要傳入額外的標(biāo)簽，可以像下面這樣在參數(shù)列表的后面加入：

channelError,?_?:=?strconv.ParseFloat(channelStatsList.Channels[i].Error,?64)
ch?<-?prometheus.MustNewConstMetric(
?messagesErrored,?prometheus.GaugeValue,?channelError,?channelName,
)

11. 在 main 函數(shù)中聲明 exporter

exporter?:=?NewExporter(mirthEndpoint,?mirthUsername,?mirthPassword)
prometheus.MustRegister(exporter)

到這里其實(shí)這個(gè) Exporter 就可以使用了，每次訪問 metrics 路由的時(shí)候，它會執(zhí)行 api 調(diào)用，并以 Prometheus Text 文本格式返回?cái)?shù)據(jù)。下面的步驟主要是方便部署了。

12. 將硬編碼的 api 路徑放到 flag 中

前面我們硬編碼了好多參數(shù)，比如應(yīng)用程序的網(wǎng)址、metrics 路由地址以及 exporter 端口，我們可以通過從命令行參數(shù)中來解析這些值使程序更加靈活。

var?(
listenAddress?=?flag.String("web.listen-address",?":9141",
?"Address?to?listen?on?for?telemetry")
metricsPath?=?flag.String("web.telemetry-path",?"/metrics",
?"Path?under?which?to?expose?metrics")
)
func?main()?{
???flag.Parse()
???...
???http.Handle(*metricsPath,?promhttp.Handler())
???log.Fatal(http.ListenAndServe(*listenAddress,?nil))
}

13. 將憑證放入環(huán)境變量

如果應(yīng)用端點(diǎn)改變了或者登錄憑證改變了怎么辦？我們可以從環(huán)境變量中來加載這些數(shù)據(jù)，在這個(gè)例子中，我們使用 godotenv 這個(gè)包來幫助將變量值存儲在本地的一個(gè)目錄中：

import?(
??"os"
)
func?main()?{
?err?:=?godotenv.Load()
?if?err?!=?nil?{
??log.Println("Error?loading?.env?file,?assume?env?variables?are?set.")
?}
?mirthEndpoint?:=?os.Getenv("MIRTH_ENDPOINT")
?mirthUsername?:=?os.Getenv("MIRTH_USERNAME")
?mirthPassword?:=?os.Getenv("MIRTH_PASSWORD")
}

整個(gè) Exporter 完整的代碼如下所示：

package?main

import?(
?"crypto/tls"
?"encoding/xml"
?"flag"
?"io/ioutil"
?"log"
?"net/http"
?"os"
?"strconv"

?"github.com/joho/godotenv"
?"github.com/prometheus/client_golang/prometheus"
?"github.com/prometheus/client_golang/prometheus/promhttp"
)

/*

*/
type?ChannelIdNameMap?struct?{
?XMLName?xml.Name???????`xml:"map"`
?Entries?[]ChannelEntry?`xml:"entry"`
}
type?ChannelEntry?struct?{
?XMLName?xml.Name?`xml:"entry"`
?Values??[]string?`xml:"string"`
}

/*

??
????c5e6a736-0e88-46a7-bf32-5b4908c4d859
????101af57f-f26c-40d3-86a3-309e74b93512
????0
????0
????0
????0
????0
??

*/
type?ChannelStatsList?struct?{
?XMLName??xml.Name???????`xml:"list"`
?Channels?[]ChannelStats?`xml:"channelStatistics"`
}
type?ChannelStats?struct?{
?XMLName???xml.Name?`xml:"channelStatistics"`
?ServerId??string???`xml:"serverId"`
?ChannelId?string???`xml:"channelId"`
?Received??string???`xml:"received"`
?Sent??????string???`xml:"sent"`
?Error?????string???`xml:"error"`
?Filtered??string???`xml:"filtered"`
?Queued????string???`xml:"queued"`
}

const?namespace?=?"mirth"
const?channelIdNameApi?=?"/api/channels/idsAndNames"
const?channelStatsApi?=?"/api/channels/statistics"

var?(
?tr?=?&http.Transport{
??TLSClientConfig:?&tls.Config{InsecureSkipVerify:?true},
?}
?client?=?&http.Client{Transport:?tr}

?listenAddress?=?flag.String("web.listen-address",?":9141",
??"Address?to?listen?on?for?telemetry")
?metricsPath?=?flag.String("web.telemetry-path",?"/metrics",
??"Path?under?which?to?expose?metrics")

?//?Metrics
?up?=?prometheus.NewDesc(
??prometheus.BuildFQName(namespace,?"",?"up"),
??"Was?the?last?Mirth?query?successful.",
??nil,?nil,
?)
?messagesReceived?=?prometheus.NewDesc(
??prometheus.BuildFQName(namespace,?"",?"messages_received_total"),
??"How?many?messages?have?been?received?(per?channel).",
??[]string{"channel"},?nil,
?)
?messagesFiltered?=?prometheus.NewDesc(
??prometheus.BuildFQName(namespace,?"",?"messages_filtered_total"),
??"How?many?messages?have?been?filtered?(per?channel).",
??[]string{"channel"},?nil,
?)
?messagesQueued?=?prometheus.NewDesc(
??prometheus.BuildFQName(namespace,?"",?"messages_queued"),
??"How?many?messages?are?currently?queued?(per?channel).",
??[]string{"channel"},?nil,
?)
?messagesSent?=?prometheus.NewDesc(
??prometheus.BuildFQName(namespace,?"",?"messages_sent_total"),
??"How?many?messages?have?been?sent?(per?channel).",
??[]string{"channel"},?nil,
?)
?messagesErrored?=?prometheus.NewDesc(
??prometheus.BuildFQName(namespace,?"",?"messages_errored_total"),
??"How?many?messages?have?errored?(per?channel).",
??[]string{"channel"},?nil,
?)
)

type?Exporter?struct?{
?mirthEndpoint,?mirthUsername,?mirthPassword?string
}

func?NewExporter(mirthEndpoint?string,?mirthUsername?string,?mirthPassword?string)?*Exporter?{
?return?&Exporter{
??mirthEndpoint:?mirthEndpoint,
??mirthUsername:?mirthUsername,
??mirthPassword:?mirthPassword,
?}
}

func?(e?*Exporter)?Describe(ch?chan<-?*prometheus.Desc)?{
?ch?<-?up
?ch?<-?messagesReceived
?ch?<-?messagesFiltered
?ch?<-?messagesQueued
?ch?<-?messagesSent
?ch?<-?messagesErrored
}

func?(e?*Exporter)?Collect(ch?chan<-?prometheus.Metric)?{
?channelIdNameMap,?err?:=?e.LoadChannelIdNameMap()
?if?err?!=?nil?{
??ch?<-?prometheus.MustNewConstMetric(
???up,?prometheus.GaugeValue,?0,
??)
??log.Println(err)
??return
?}
?ch?<-?prometheus.MustNewConstMetric(
??up,?prometheus.GaugeValue,?1,
?)

?e.HitMirthRestApisAndUpdateMetrics(channelIdNameMap,?ch)
}

func?(e?*Exporter)?LoadChannelIdNameMap()?(map[string]string,?error)?{
?//?Create?the?map?of?channel?id?to?names
?channelIdNameMap?:=?make(map[string]string)

?req,?err?:=?http.NewRequest("GET",?e.mirthEndpoint+channelIdNameApi,?nil)
?if?err?!=?nil?{
??return?nil,?err
?}

?//?This?one?line?implements?the?authentication?required?for?the?task.
?req.SetBasicAuth(e.mirthUsername,?e.mirthPassword)
?//?Make?request?and?show?output.
?resp,?err?:=?client.Do(req)
?if?err?!=?nil?{
??return?nil,?err
?}

?body,?err?:=?ioutil.ReadAll(resp.Body)
?resp.Body.Close()
?if?err?!=?nil?{
??return?nil,?err
?}
?//?fmt.Println(string(body))

?//?we?initialize?our?array
?var?channelIdNameMapXML?ChannelIdNameMap
?//?we?unmarshal?our?byteArray?which?contains?our
?//?xmlFiles?content?into?'users'?which?we?defined?above
?err?=?xml.Unmarshal(body,?&channelIdNameMapXML)
?if?err?!=?nil?{
??return?nil,?err
?}

?for?i?:=?0;?i?len(channelIdNameMapXML.Entries);?i++?{
??channelIdNameMap[channelIdNameMapXML.Entries[i].Values[0]]?=?channelIdNameMapXML.Entries[i].Values[1]
?}

?return?channelIdNameMap,?nil
}

func?(e?*Exporter)?HitMirthRestApisAndUpdateMetrics(channelIdNameMap?map[string]string,?ch?chan<-?prometheus.Metric)?{
?//?Load?channel?stats
?req,?err?:=?http.NewRequest("GET",?e.mirthEndpoint+channelStatsApi,?nil)
?if?err?!=?nil?{
??log.Fatal(err)
?}

?//?This?one?line?implements?the?authentication?required?for?the?task.
?req.SetBasicAuth(e.mirthUsername,?e.mirthPassword)
?//?Make?request?and?show?output.
?resp,?err?:=?client.Do(req)
?if?err?!=?nil?{
??log.Fatal(err)
?}

?body,?err?:=?ioutil.ReadAll(resp.Body)
?resp.Body.Close()
?if?err?!=?nil?{
??log.Fatal(err)
?}
?//?fmt.Println(string(body))

?//?we?initialize?our?array
?var?channelStatsList?ChannelStatsList
?//?we?unmarshal?our?byteArray?which?contains?our
?//?xmlFiles?content?into?'users'?which?we?defined?above
?err?=?xml.Unmarshal(body,?&channelStatsList)
?if?err?!=?nil?{
??log.Fatal(err)
?}

?for?i?:=?0;?i?len(channelStatsList.Channels);?i++?{
??channelName?:=?channelIdNameMap[channelStatsList.Channels[i].ChannelId]

??channelReceived,?_?:=?strconv.ParseFloat(channelStatsList.Channels[i].Received,?64)
??ch?<-?prometheus.MustNewConstMetric(
???messagesReceived,?prometheus.GaugeValue,?channelReceived,?channelName,
??)

??channelSent,?_?:=?strconv.ParseFloat(channelStatsList.Channels[i].Sent,?64)
??ch?<-?prometheus.MustNewConstMetric(
???messagesSent,?prometheus.GaugeValue,?channelSent,?channelName,
??)

??channelError,?_?:=?strconv.ParseFloat(channelStatsList.Channels[i].Error,?64)
??ch?<-?prometheus.MustNewConstMetric(
???messagesErrored,?prometheus.GaugeValue,?channelError,?channelName,
??)

??channelFiltered,?_?:=?strconv.ParseFloat(channelStatsList.Channels[i].Filtered,?64)
??ch?<-?prometheus.MustNewConstMetric(
???messagesFiltered,?prometheus.GaugeValue,?channelFiltered,?channelName,
??)

??channelQueued,?_?:=?strconv.ParseFloat(channelStatsList.Channels[i].Queued,?64)
??ch?<-?prometheus.MustNewConstMetric(
???messagesQueued,?prometheus.GaugeValue,?channelQueued,?channelName,
??)
?}

?log.Println("Endpoint?scraped")
}

func?main()?{
?err?:=?godotenv.Load()
?if?err?!=?nil?{
??log.Println("Error?loading?.env?file,?assume?env?variables?are?set.")
?}

?flag.Parse()

?mirthEndpoint?:=?os.Getenv("MIRTH_ENDPOINT")
?mirthUsername?:=?os.Getenv("MIRTH_USERNAME")
?mirthPassword?:=?os.Getenv("MIRTH_PASSWORD")

?exporter?:=?NewExporter(mirthEndpoint,?mirthUsername,?mirthPassword)
?prometheus.MustRegister(exporter)

?http.Handle(*metricsPath,?promhttp.Handler())
?http.HandleFunc("/",?func(w?http.ResponseWriter,?r?*http.Request)?{
??w.Write([]byte(`
?????????????
?????????????
?????????????
Mirth?Channel?Exporter

?????????????
Metrics

?????????????
?????????????`))
?})
?log.Fatal(http.ListenAndServe(*listenAddress,?nil))
}

14. 編寫一個(gè) Makefile 文件，方便在不同平臺上快速構(gòu)建

Makefile 可以讓你在開發(fā)過程中省去很多多余的操作，比如我們要構(gòu)建多個(gè)平臺的構(gòu)建程序，可以創(chuàng)建如下所示的 Makefile 文件。

linux:
???GOOS=linux?GOARCH=amd64?go?build
mac:
???GOOS=darwin?GOARCH=amd64?go?build

只要調(diào)用 make mac 或 make linux 命令就可以看到不同的可執(zhí)行文件出現(xiàn)。

15. 編寫一個(gè) service 文件，將這個(gè) go 應(yīng)用作為守護(hù)進(jìn)程運(yùn)行

我們可以為這個(gè) Exporter 編寫一個(gè) service 文件或者 Dockerfile 文件來管理該應(yīng)用，比如這里我們直接在 Centos 7 上用 systemd 來管理該應(yīng)用。這可以編寫一個(gè)如下所示的 service 文件：

[Unit]
Description=mirth?channel?exporter
After=network.target
StartLimitIntervalSec=0
[Service]
Type=simple
Restart=always
RestartSec=1
WorkingDirectory=/mirth/mirthconnect
EnvironmentFile=/etc/sysconfig/mirth_channel_exporter
ExecStart=/mirth/mirthconnect/mirth_channel_exporter

[Install]
WantedBy=multi-user.target

到這里就完成了使用 Golang 編寫一個(gè)簡單的 Prometheus Exporter。

“

原文鏈接：https://medium.com/teamzerolabs/15-steps-to-write-an-application-prometheus-exporter-in-go-9746b4520e26

”