|NO.Z.00062|——————————|^^ 部署 ^^|——|Hadoop&实时数仓.V02|——|可视化.v02|监控：Prometheus部署.V2|

一、prometheus部署：安装go语言环境

### --- 下载版本包并解压

~~~     # 由于Prometheus 是用golang开发的，所以首先安装一个go环境，Go语言是跨平台，支持Windows、Linux、
~~~     # Windows ：go1.8.3.windows-amd64.msi (78MB)
~~~     Linux ：      go1.8.3.linux-amd64.tar.gz (86MB)
~~~     Mac ：        go1.8.3.darwin-amd64.tar.gz (85MB)
~~~     Source：  go1.8.3.src.tar.gz (15MB)

~~~     # 下载go语言版本包：https://storage.googleapis.com/golang/go1.8.3.linux-amd64.tar.gz
[root@hadoop00 software]# wget -c https://storage.googleapis.com/golang/go1.8.3.linux-amd64.tar.gz

~~~     # 解压go语言版本包
[root@hadoop00 software]# tar -zxvf go1.8.3.linux-amd64.tar.gz -C ../servers/

### --- 配置go语言环境变量

~~~     # 配置go语言环境变量
[root@hadoop00 ~]# vim /etc/profile
## GO_HOME
export GO_HOME=/opt/yanqi/servers/go
export PATH=:$GO_HOME/bin:$PATH
export GOROOT=/opt/yanqi/servers/go
export PATH=:$GOROOT/bin:$PATH

~~~     # 使环境变量生效

[root@hadoop00 ~]# source /etc/profile

### --- 验证go语言环境是否生效

~~~     # 验证go语言环境版本
[root@hadoop00 ~]# go version
go version go1.8.3 linux/amd64

二、prometheus部署

### --- 下载prometheus版本包

~~~     # 下载prometheus版本包
[root@hadoop00 software]# wget -c https://github.com/prometheus/prometheus/releases/download/v2.22.1/prometheus-2.22.1.linux-amd64.tar.gz

### --- 解压prometheus版本包

~~~     # 加压prometheus版本包
[root@hadoop00 software]# tar -zxvf prometheus-2.22.1.linux-amd64.tar.gz -C ../servers/
~~~     # 修改版本包名称
[root@hadoop00 ~]# cd /opt/yanqi/servers/
[root@hadoop00 servers]# mv prometheus-2.22.1.linux-amd64/ prometheus

~~~     # 修改目录权限

[root@hadoop00 ~]# chown -R root:root /opt/yanqi/servers/prometheus/

### --- 修改prometheus配置文件

~~~     # 修改prometheus版本包名称
[root@hadoop00 ~]# vim /opt/yanqi/servers/prometheus/prometheus.yml
 ~~~    # 第21~45行：配置如下参数
 scrape_configs:
  # The job name is added as a label `job=<job_name>` to any timeseries scraped from this config.
  - job_name: 'prometheus'

    # metrics_path defaults to '/metrics'
    # scheme defaults to 'http'.

    static_configs:
    - targets: ['hadoop00:9090']

  - job_name: 'bigdata-hadoop00'
    static_configs:
    - targets: ['hadoop00:9100']

  - job_name: 'bigdata-hadoop01'
    static_configs:
    - targets: ['hadoop01:9100']

  - job_name: 'bigdata-hadoop02'
    static_configs:
    - targets: ['hadoop02:9100']

  - job_name: 'bigdata-hadoop03'
    static_configs:
    - targets: ['hadoop03:9100']

  - job_name: 'bigdata-pushgateway'
    static_configs:
    - targets: ['hadoop00:9091']

### --- 启动prometheus服务

~~~     # 启动prometheus服务：启动之前同步时间：ntpdate -u ntp.api.bz
[root@hadoop00 ~]# /opt/yanqi/servers/prometheus/prometheus  --config.file=/opt/yanqi/servers/prometheus/prometheus.yml

三、通过web_UI访问prometheus：

### --- 通过web_UI访问prometheus：

~~~     # 1、通过web_UI访问prometheus：http://hadoop00:9090/graph
~~~     # 2、查看监控到的主机

附录一：报错处理一：

### --- 报错现象：

[root@hadoop00 ~]# /opt/yanqi/servers/prometheus/prometheus  --config.file=/opt/yanqi/servers/prometheus/prometheus.yml
~~~输出报错信息:
level=warn ts=2021-11-30T08:50:49.794Z caller=scrape.go:1091 component="scrape manager" scrape_pool=bigdata-pushgateway target=http://hadoop00:9091/metrics msg="Appending scrape report failed" err="out of bounds"
level=warn ts=2021-11-30T08:50:50.466Z caller=scrape.go:1091 component="scrape manager" scrape_pool=bigdata-hadoop03 target=http://hadoop03:9100/metrics msg="Appending scrape report failed" err="out of bounds"
level=warn ts=2021-11-30T08:50:51.664Z caller=scrape.go:1091 component="scrape manager" scrape_pool=bigdata-hadoop01 target=http://hadoop01:9100/metrics msg="Appending scrape report failed" err="out of bounds"
level=warn ts=2021-11-30T08:50:53.331Z caller=scrape.go:1091 component="scrape manager" scrape_pool=bigdata-hadoop02 target=http://hadoop02:9100/metrics msg="Appending scrape report failed" err="out of bounds"
level=warn ts=2021-11-30T08:52:54.730Z caller=scrape.go:1091 component="scrape manager" scrape_pool=prometheus target=http://hadoop00:9090/metrics msg="Appending scrape report failed" err="out of bounds"

### --- 报错分析：

~~~     # 是时间跳变导致的，与Prometheus 机制有关;prometheus和grafana都会获取不到数据
~~~     # 猜测如下，vmotion后操作系统时间发生了向后跳变，vmotion从跳变时间开始取数，grafna当前当前时间显示无数据，修改操作系统时间后与Prometheus最后取数时间发生冲突，报“Error on ingesting samples that are too old or are too far into the future”无法取数。
~~~     # 解决方法：清空prometheus2.0.0.data.metrics 文件后重启prometheus 数据正常。

### --- 解决方案：
~~~     # 删除prometheus下的data.metrics目录，并重启prometheus服务

[root@hadoop00 ~]# rm -rf /opt/yanqi/servers/prometheus/data/

Walter Savage Landor:strove with none,for none was worth my strife.Nature I loved and, next to Nature, Art:I warm'd both hands before the fire of life.It sinks, and I am ready to depart

——W.S.Landor