阅读 143

Kubernetes中Nginx配置热加载的全过程

Nginx已经是互联网IT业界一个无敌的存在,作为反向代理、负载均衡、Web服务器等多种角色的扮演者,下面这篇文章主要给大家介绍了关于Kubernetes中Nginx配置热加载的相关资料,需要的朋友可以参考下

目录

  • 前言

  • 使用方法

  • 总结

前言

Nginx本身是支持热更新的，通过nginx -s reload指令，实际通过向进程发送HUB信号实现不停服重新加载配置，然而在Docker或者Kubernetes中，每次都需要进容器执行nginx -s reload指令，单docker容器还好说，可以在外面通过exec指定容器执行该指令进行热加载，Kubernetes的话，就比较难受了

今天介绍一下Kubernetes中Nginx热加载配置的处理方法——reloader

reloader地址：https://github.com/stakater/Reloader

reloader主要就是用来监测ConfigMap或Secret的变化，然后对相关DeploymentConfig的Deployment、DaemonSet执行滚动升级

reloader需要kubernetes1.9以上的版本才支持

使用方法

首先是安装部署reloader

1
2
# 直接通过官方yaml文件部署
kubectl apply -f https://raw.githubusercontent.com/stakater/Reloader/master/deployments/kubernetes/reloader.yaml

默认情况下reloader是部署在default命名空间，但是它是监控所有命名空间的configmaps和secrets

当然，如果不想监控某个configmap或secret，可以通过--resources-to-ignore=configMaps/secrets来忽略某个资源

部署成功后，就可以直接使用了，我提前部署了nginx和configmap

这是目前的配置，看一下Nginx目前的配置

接着，我修改Nginx的Deployment，添加reloader，监听nginx-config这个ConfigMap，执行reload

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
{
  "kind": "Deployment",
  "apiVersion": "extensions/v1beta1",
  "metadata": {
    "name": "nginx",
    "namespace": "default",
    "selfLink": "/apis/extensions/v1beta1/namespaces/default/deployments/nginx",
    "uid": "7eee5fa8-7514-11ec-a916-0210d5e9ca3b",
    "resourceVersion": "286141",
    "generation": 10,
    "creationTimestamp": "2022-01-14T08:32:23Z",
    "labels": {
      "k8s-app": "nginx"
    },
    "annotations": {
      "deployment.kubernetes.io/revision": "9",
      "description": "nginx应用"
      # 主要是这行
      "reloader.stakater.com/reload": "nginx-config"
    }
  },
  "spec": {
    "replicas": 1,
    "selector": {
      "matchLabels": {
        "k8s-app": "nginx"
      }
    }
    ……

然后apply该Deployment，之后我们去更新ConfigMap，更新nginx配置文件

更新完成，去掉proxy_redirect，然后去看nginx容器是否执行滚动更新

可以看到，nginx执行了滚动更新，接着看下nginx配置文件是否更新

这样很简单的通过reloader就可以实现Nginx的配置热加载

除了这种方法，常见的方法还有使用sidecar，通过sidecar去做的话，需要自己写监听脚本，比较麻烦，但是有时候也相对灵活，这里也附一个sidecar的python脚本

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
#!/usr/bin/env python
# -*- encoding: utf8 -*-
"""
需求：nginx配置文件变化，自动更新配置文件，类似nginx -s reload
实现：
    1、用pyinotify实时监控nginx配置文件变化
    2、如果配置文件变化，给系统发送HUP来reload nginx
"""
import os
import re
import pyinotify
import logging
from threading import Timer
 
# Param
LOG_PATH = "/root/python/log"
CONF_PATHS = [
  "/etc/nginx",
]
DELAY = 5
SUDO = False
RELOAD_COMMAND = "nginx -s reload"
if SUDO:
  RELOAD_COMMAND = "sudo " + RELOAD_COMMAND
 
# Log
logger = logging.getLogger(__name__)
logger.setLevel(level = logging.INFO)
log_handler = logging.FileHandler(LOG_PATH)
log_handler.setLevel(logging.INFO)
log_formatter = logging.Formatter('%(asctime)s - %(levelname)s - %(message)s')
log_handler.setFormatter(log_formatter)
logger.addHandler(log_handler)
 
# Reloader
def reload_nginx():
  os.system(RELOAD_COMMAND)
  logger.info("nginx is reloaded")
 
t = Timer(DELAY, reload_nginx)
 
def trigger_reload_nginx(pathname, action):
  logger.info("nginx monitor is triggered because %s is %s" % (pathname, action))
  global t
  if t.is_alive():
    t.cancel()
    t = Timer(DELAY, reload_nginx)
    t.start()
  else:
    t = Timer(DELAY, reload_nginx)
    t.start()
 
events = pyinotify.IN_MODIFY | pyinotify.IN_CREATE | pyinotify.IN_DELETE
 
watcher = pyinotify.WatchManager()
watcher.add_watch(CONF_PATHS, events, rec=True, auto_add=True)
 
class EventHandler(pyinotify.ProcessEvent):
  def process_default(self, event):
    if event.name.endswith(".conf"):
      if event.mask == pyinotify.IN_CREATE:
        action = "created"
      if event.mask == pyinotify.IN_MODIFY:
        action = "modified"
      if event.mask == pyinotify.IN_DELETE:
        action = "deleted"
      trigger_reload_nginx(event.pathname, action)
 
handler = EventHandler()
notifier = pyinotify.Notifier(watcher, handler)
 
# Start
logger.info("Start Monitoring")
notifier.loop()

如果喜欢用go的，这里也提供go脚本

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
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
package main
 
import (
    "log"
    "os"
    "path/filepath"
    "syscall"
 
    "github.com/fsnotify/fsnotify"
    proc "github.com/shirou/gopsutil/process"
)
 
const (
    nginxProcessName = "nginx"
    defaultNginxConfPath = "/etc/nginx"
    watchPathEnvVarName = "WATCH_NGINX_CONF_PATH"
)
 
var stderrLogger = log.New(os.Stderr, "error: ", log.Lshortfile)
var stdoutLogger = log.New(os.Stdout, "", log.Lshortfile)
 
func getMasterNginxPid() (int, error) {
    processes, processesErr := proc.Processes()
    if processesErr != nil {
        return 0, processesErr
    }
 
    nginxProcesses := map[int32]int32{}
 
    for _, process := range processes {
        processName, processNameErr := process.Name()
        if processNameErr != nil {
            return 0, processNameErr
        }
 
        if processName == nginxProcessName {
            ppid, ppidErr := process.Ppid()
 
            if ppidErr != nil {
                return 0, ppidErr
            }
 
            nginxProcesses[process.Pid] = ppid
        }
    }
 
    var masterNginxPid int32
 
    for pid, ppid := range nginxProcesses {
        if ppid == 0 {
            masterNginxPid = pid
 
            break
        }
    }
 
    stdoutLogger.Println("found master nginx pid:", masterNginxPid)
 
    return int(masterNginxPid), nil
}
 
func signalNginxReload(pid int) error {
    stdoutLogger.Printf("signaling master nginx process (pid: %d) -> SIGHUP\n", pid)
    nginxProcess, nginxProcessErr := os.FindProcess(pid)
 
    if nginxProcessErr != nil {
        return nginxProcessErr
    }
 
    return nginxProcess.Signal(syscall.SIGHUP)
}
 
func main() {
    watcher, watcherErr := fsnotify.NewWatcher()
    if watcherErr != nil {
        stderrLogger.Fatal(watcherErr)
    }
    defer watcher.Close()
 
    done := make(chan bool)
    go func() {
        for {
            select {
            case event, ok := <-watcher.Events:
                if !ok {
                    return
                }
 
                if event.Op&fsnotify.Create == fsnotify.Create {
                    if filepath.Base(event.Name) == "..data" {
                        stdoutLogger.Println("config map updated")
 
                        nginxPid, nginxPidErr := getMasterNginxPid()
                        if nginxPidErr != nil {
                            stderrLogger.Printf("getting master nginx pid failed: %s", nginxPidErr.Error())
 
                            continue
                        }
 
                        if err := signalNginxReload(nginxPid); err != nil {
                            stderrLogger.Printf("signaling master nginx process failed: %s", err)
                        }
                    }
                }
            case err, ok := <-watcher.Errors:
                if !ok {
                    return
                }
                stderrLogger.Printf("received watcher.Error: %s", err)
            }
        }
    }()
 
    pathToWatch, ok := os.LookupEnv(watchPathEnvVarName)
    if !ok {
        pathToWatch = defaultNginxConfPath
    }
 
    stdoutLogger.Printf("adding path: `%s` to watch\n", pathToWatch)
 
    if err := watcher.Add(pathToWatch); err != nil {
        stderrLogger.Fatal(err)
    }
    <-done
}

ok,今天的内容就到这里

总结

到此这篇关于Kubernetes中Nginx配置热加载的文章就介绍到这了

原文链接：https://mp.weixin.qq.com/s/dcqprANA0m6gGEvnF2l4uQ


文章分类
代码人生
文章标签
kubernetes
nginx
过程
版权声明：本站是系统测试站点，无实际运营。本文内容由互联网用户自发贡献，该文观点仅代表作者本人。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容， 请发送邮件至 XXXXXXo@163.com 举报，一经查实，本站将立刻删除。
相关推荐