自从Go 1.5开始,每次Go release, Gopher Brian Hatfield都会将自己对新版Go的runtime的性能数据(与之前Go版本的比较)在twitter上晒出来。就连Go team staff在世界各地做speaking时也在slide中引用Brian的图片。后来,Brian Hatfield将其用于度量runtime性能数据的代码打包成library并放在github上开源了,我们也可以使用这个library来建立我们自己的Go Runtime metrics设施了。这里简要说一下搭建的步骤。

一、环境与原理

Brian Hatfield的go-runtime-metrics library实现的很简单,其runtime data来自于Go runtime package中的MemStats、NumGoroutine和NumCgoCall等。被测试目标程序只需要import该library即可输出runtime states数据:

import _ "github.com/bmhatfield/go-runtime-metrics"

go-runtime-metrics library将启动一个单独的goroutine,并定时上报runtime数据。目前该library仅支持向statsD输出数据,用户可以通过配置将statsD的数据导入graphite并使用graphite web查看,流程如下图:

img{512x368}

本次实验环境为ubuntu 16.04.1:

$ uname -rmn
tonybai-ThinkCentre-M6600s-N000 4.4.0-83-generic x86_64

二、搭建步骤

1、安装go-runtime-metrics library

我们直接go get就可以下载go-runtime-metrics library:

$ go get github.com/bmhatfield/go-runtime-metrics

我们编写一个目标程序:

//main.go
package main

import (
    "flag"
    "log"
    "net/http"
    "os"

    _ "github.com/bmhatfield/go-runtime-metrics"
)

func main() {
    flag.Parse()

    cwd, err := os.Getwd()
    if err != nil {
        log.Fatal(err)
    }

    srv := &http.Server{
        Addr:    ":8000", // Normally ":443"
        Handler: http.FileServer(http.Dir(cwd)),
    }
    log.Fatal(srv.ListenAndServe())
}

我的ubuntu主机上安装了四个go版本,它们分别是go 1.5.4、go 1.7.6、go 1.8.3和go1.9beta2,于是我们分别用这四个版本的server作为被测程序进行go runtime数据上报,以便对比。

$ GOROOT=~/.bin/go154 ~/.bin/go154/bin/go build -o server-go154 main.go
$ GOROOT=~/.bin/go174 ~/.bin/go174/bin/go build -o server-go174 main.go
$ GOROOT=~/.bin/go183 ~/.bin/go183/bin/go build -o server-go183 main.go
$ GOROOT=~/.bin/go19beta2 ~/.bin/go19beta2/bin/go build -o server-go19beta2 main.go

$ ls -l

-rwxr-xr-x 1 tonybai tonybai 6861176 7月   4 13:49 server-go154
-rwxrwxr-x 1 tonybai tonybai 5901876 7月   4 13:50 server-go174
-rwxrwxr-x 1 tonybai tonybai 6102879 7月   4 13:51 server-go183
-rwxrwxr-x 1 tonybai tonybai 6365648 7月   4 13:51 server-go19beta2

2、安装、配置和运行statsD

statsD这个工具用于收集统计信息,并将聚合后的信息发给后端服务(比如:graphite)。statsD是采用js实现的服务,因此需要安装nodejsnpm和相关modules:

$ sudo apt-get install nodejs
$ sudo apt-get install npm

接下来,我们将statsD项目clone到本地并根据exampleConfig.js模板配置一个我们自己用的goruntimemetricConfig.js(基本上就是保留默认配置):

// goruntimemetricConfig.js
{
  graphitePort: 2003
, graphiteHost: "127.0.0.1"
, port: 8125
, backends: [ "./backends/graphite" ]
}

启动statsD:

$ nodejs stats.js goruntimemetricConfig.js
3 Jul 11:14:20 - [7939] reading config file: goruntimemetricConfig.js
3 Jul 11:14:20 - server is up INFO

启动成功!

3、安装、配置和运行graphite

graphite是一种存储时序监控数据,并可以按用户需求以图形化形式展示数据的工具,它包括三个组件:

whisper是一种基于file的时序数据库格式,同时whisper也提供了相应的命令和API供其他组件调用以操作时序数据库;

carbon用于读取外部推送的metrics信息,进行聚合并写入db,它还支持缓存热点数据,提升访问效率。

graphite-web则是针对用户的图形化系统,用于定制展示监控数据的。

Graphite的安装和配置是略微繁琐的,我们一步一步慢慢来。

a) 安装graphite

$sudo apt-get install graphite-web graphite-carbon

whisper将作为依赖自动被安装。

b) local_settings.py

graphite的主配置文件在/etc/graphite/local_settings.py,文件里面有很多配置项,这里仅列出有关的,且本次生效的配置:

// /etc/graphite/local_settings.py

TIME_ZONE = 'Asia/Shanghai'

LOG_RENDERING_PERFORMANCE = True
LOG_CACHE_PERFORMANCE = True
LOG_METRIC_ACCESS = True

GRAPHITE_ROOT = '/usr/share/graphite-web'

CONF_DIR = '/etc/graphite'
STORAGE_DIR = '/var/lib/graphite/whisper'
CONTENT_DIR = '/usr/share/graphite-web/static'

WHISPER_DIR = '/var/lib/graphite/whisper'
LOG_DIR = '/var/log/graphite'
INDEX_FILE = '/var/lib/graphite/search_index'  # Search index file

DATABASES = {
    'default': {
        'NAME': '/var/lib/graphite/graphite.db',
        'ENGINE': 'django.db.backends.sqlite3',
        'USER': '',
        'PASSWORD': '',
        'HOST': '',
        'PORT': ''
    }
}

c) 同步数据库

接下来执行下面两个命令来做database sync(同步):

$ sudo graphite-manage migrate auth
.. ....
Operations to perform:
  Apply all migrations: auth
Running migrations:
  Rendering model states... DONE
  Applying contenttypes.0001_initial... OK
  Applying contenttypes.0002_remove_content_type_name... OK
  Applying auth.0001_initial... OK
  Applying auth.0002_alter_permission_name_max_length... OK
  Applying auth.0003_alter_user_email_max_length... OK
  Applying auth.0004_alter_user_username_opts... OK
  Applying auth.0005_alter_user_last_login_null... OK
  Applying auth.0006_require_contenttypes_0002... OK

$ sudo graphite-manage syncdb

Operations to perform:
  Synchronize unmigrated apps: account, cli, render, whitelist, metrics, url_shortener, dashboard, composer, events, browser
  Apply all migrations: admin, contenttypes, tagging, auth, sessions
Synchronizing apps without migrations:
  Creating tables...
    Creating table account_profile
    Creating table account_variable
    Creating table account_view
    Creating table account_window
    Creating table account_mygraph
    Creating table dashboard_dashboard
    Creating table events_event
    Creating table url_shortener_link
    Running deferred SQL...
  Installing custom SQL...
Running migrations:
  Rendering model states... DONE
  Applying admin.0001_initial... OK
  Applying sessions.0001_initial... OK
  Applying tagging.0001_initial... OK

You have installed Django's auth system, and don't have any superusers defined.
Would you like to create one now? (yes/no): yes
Username (leave blank to use 'root'):
Email address: xx@yy.com
Password:
Password (again):
Superuser created successfully.

这里我们创建一个superuser:root,用于登录graphite-web时使用。

d) 配置carbon

涉及carbon的配置文件如下,我们保持默认配置不动:

/etc/carbon/carbon.conf(内容太多,这里不列出来了)

/etc/carbon/storage-schemas.conf
[carbon]
pattern = ^carbon\.
retentions = 60:90d

[default_1min_for_1day]
pattern = .*
retentions = 60s:1d

[stats]
pattern = ^stats.*
retentions = 10s:6h,1min:6d,10min:1800d

carbon有一个cache功能,我们通过下面步骤可以将其打开:

打开carbon-cache使能开关:

$ vi /etc/default/graphite-carbon
CARBON_CACHE_ENABLED=true

启动carbon-cache:

$ sudo cp /usr/share/doc/graphite-carbon/examples/storage-aggregation.conf.example /etc/carbon/storage-aggregation.conf
$ systemctl start carbon-cache

e) 启动graphite-web

graphite-web支持多种主流web server,这里以apache2为例,graphite-web将mod-wsgi方式部署在apache2下面:

$sudo apt-get install apache2 libapache2-mod-wsgi

$ sudo service apache2 start

$ sudo a2dissite 000-default
Site 000-default disabled.

$ sudo service apache2 reload

$ sudo cp /usr/share/graphite-web/apache2-graphite.conf /etc/apache2/sites-available

$ sudo  a2ensite apache2-graphite
Enabling site apache2-graphite.
To activate the new configuration, you need to run:
  service apache2 reload

$ sudo systemctl reload apache2

由于apache2的Worker process默认以www-data:www-data用户权限运行,但数据库文件的访问权限却是:_graphite:_graphite:

$ ll /var/lib/graphite/graphite.db
-rw-r--r-- 1 _graphite _graphite 72704 7月   3 13:48 /var/lib/graphite/graphite.db

我们需要修改一下apache worker的user:

$ sudo vi /etc/apache2/envvars

export APACHE_RUN_USER=_graphite
export APACHE_RUN_GROUP=_graphite

重启apache2生效!使用Browser打开:http://127.0.0.1,如无意外,你将看到下面graphite-web的首页:

img{512x368}

三、执行benchmarking

这里我将使用wrk这个http benchmarking tool分别对前面的四个版本的目标程序(server-go154 server-go174 server-go183 server-go19beta2)进行benchmarking test,每个目标程序接收10分钟的请求:

$ ./server-go154
$ wrk -t12 -c400 -d10m http://127.0.0.1:8000

$ ./server-go174
$ wrk -t12 -c400 -d10m http://127.0.0.1:8000

$ ./server-go183
$ wrk -t12 -c400 -d10m http://127.0.0.1:8000

$ ./server-go19beta2
$ wrk -t12 -c400 -d10m http://127.0.0.1:8000

四、结果展示

用浏览器打开graphite-web,在左边的tree标签下以此打开树形结构:Metrics -> stats -> gauges -> go -> YOUR_HOST_NAME -> mem -> gc -> pause,如果顺利的话,你将会在Graphite Composer窗口看到折线图,我们也以GC pause为例,GC pause也是gopher们最为关心的:

img{512x368}

通过这幅图(左侧坐标轴的单位为nanoseconds),我们大致可以看出:

Go 1.5.4的GC pause约在600μs左右;
Go 1.7.4的GC pause约在300μs左右;
Go 1.8.3和Go 1.9beta2的GC pause基本都在100μs以下了。Go 1.9的GC改进似乎不大。不过这里我的程序也并不足够典型。

其他结果:

Go routines number:

img{512x368}

GC count:

img{512x368}

memory allocations:

img{512x368}

除了查看单个指标曲线,你也可以通过graphite-web提供的dashboard功能定制你要monitor的面板,这里就不赘述了。

五、参考资料


微博:@tonybai_cn
微信公众号:iamtonybai
github.com: https://github.com/bigwhite

© 2017, bigwhite. 版权所有.

Related posts:

  1. 制作go-talks.appspot.com应用镜像
  2. Go 1.6中值得关注的几个变化
  3. Go 1.5中值得关注的几个变化
  4. Go 1.7中值得关注的几个变化
  5. Caddy,一个用Go实现的Web Server