Linux 高可用(HA)集羣之keepalived詳解

1、前言html

2、Keepalived 詳解前端

3、環境準備node

4、LVS+Keepalived 實現高可用的前端負載均衡器vim

 

1、前言後端

       這篇文章是前幾篇文章的總結,咱們先簡單的總結一下咱們前面講解的內容,前面咱們講解了,LVS(負載均衡器)、Heartbeat、Corosync、Pacemaker、Web高可用集羣、MySQL高可用集羣、DRDB、iscsi、gfs二、cLVM等,惟一沒有講解的就是LVS可用,也就是前端高可用,咱們這一篇博文主要講解內容。在說這個以前咱們得和你們討論一個問題,也是好多博友問的問題。Heartbeat、Corosync、Keepalived這三個集羣組件咱們到底選哪一個好,首先我想說明的是,Heartbeat、Corosync是屬於同一類型,Keepalived與Heartbeat、Corosync,根本不是同一類型的。Keepalived使用的vrrp協議方式,虛擬路由冗餘協議 (Virtual Router Redundancy Protocol,簡稱VRRP);Heartbeat或Corosync是基於主機或網絡服務的高可用方式;簡單的說就是,Keepalived的目的是模擬路由器的高可用,Heartbeat或Corosync的目的是實現Service的高可用。因此通常Keepalived是實現前端高可用,經常使用的前端高可用的組合有,就是咱們常見的LVS+Keepalived、Nginx+Keepalived、HAproxy+Keepalived。而Heartbeat或Corosync是實現服務的高可用,常見的組合有Heartbeat v3(Corosync)+Pacemaker+NFS+Httpd 實現Web服務器的高可用、Heartbeat v3(Corosync)+Pacemaker+NFS+MySQL 實現MySQL服務器的高可用。總結一下,Keepalived中實現輕量級的高可用,通常用於前端高可用,且不須要共享存儲,通常經常使用於兩個節點的高可用。而Heartbeat(或Corosync)通常用於服務的高可用,且須要共享存儲,通常用於多節點的高可用。這個問題咱們說明白了,又有博友會問了,那heartbaet與corosync咱們又應該選擇哪一個好啊,我想說咱們通常用corosync,由於corosync的運行機制更優於heartbeat,就連從heartbeat分離出來的pacemaker都說在之後的開發當中更傾向於corosync,因此如今corosync+pacemaker是最佳組合。但說實話我對於軟件沒有任何傾向性,因此我把全部的集羣軟件都和你們說了一下,我認爲無論什麼軟件,只要它能存活下來都有它的特色和應用領域,只有把特定的軟件放在特定的位置才能發揮最大的做用,那首先咱們得對這個軟件有全部瞭解。學習一種軟件的最好方法,就是去查官方文檔。好了說了那麼多但願你們有所收穫,下面咱們來講一說keepalived。centos

2、Keepalived 詳解安全

1.Keepalived 定義 bash

       Keepalived 是一個基於VRRP協議來實現的LVS服務高可用方案,能夠利用其來避免單點故障。一個LVS服務會有2臺服務器運行Keepalived,一臺爲主服務器(MASTER),一臺爲備份服務器(BACKUP),可是對外表現爲一個虛擬IP,主服務器會發送特定的消息給備份服務器,當備份服務器收不到這個消息的時候,即主服務器宕機的時候, 備份服務器就會接管虛擬IP,繼續提供服務,從而保證了高可用性。Keepalived是VRRP的完美實現,所以在介紹keepalived以前,先介紹一下VRRP的原理。服務器

2.VRRP 協議簡介網絡

在現實的網絡環境中,兩臺須要通訊的主機大多數狀況下並無直接的物理鏈接。對於這樣的狀況,它們之間路由怎樣選擇?主機如何選定到達目的主機的下一跳路由,這個問題一般的解決方法有二種:

  • 在主機上使用動態路由協議(RIP、OSPF等)

  • 在主機上配置靜態路由

很明顯,在主機上配置動態路由是很是不切實際的,由於管理、維護成本以及是否支持等諸多問題。配置靜態路由就變得十分流行,但路由器(或者說默認網關default gateway)卻常常成爲單點故障。VRRP的目的就是爲了解決靜態路由單點故障問題,VRRP經過一競選(election)協議來動態的將路由任務交給LAN中虛擬路由器中的某臺VRRP路由器。

3.VRRP 工做機制

       在一個VRRP虛擬路由器中,有多臺物理的VRRP路由器,可是這多臺的物理的機器並不能同時工做,而是由一臺稱爲MASTER的負責路由工做,其它的都是BACKUP,MASTER並不是一成不變,VRRP讓每一個VRRP路由器參與競選,最終獲勝的就是MASTER。MASTER擁有一些特權,好比,擁有虛擬路由器的IP地址,咱們的主機就是用這個IP地址做爲靜態路由的。擁有特權的MASTER要負責轉發發送給網關地址的包和響應ARP請求。

       VRRP經過競選協議來實現虛擬路由器的功能,全部的協議報文都是經過IP多播(multicast)包(多播地址224.0.0.18)形式發送的。虛擬路由器由VRID(範圍0-255)和一組IP地址組成,對外表現爲一個周知的MAC地址。因此,在一個虛擬路由 器中,無論誰是MASTER,對外都是相同的MAC和IP(稱之爲VIP)。客戶端主機並不須要由於MASTER的改變而修改本身的路由配置,對客戶端來講,這種主從的切換是透明的。

       在一個虛擬路由器中,只有做爲MASTER的VRRP路由器會一直髮送VRRP通告信息(VRRPAdvertisement message),BACKUP不會搶佔MASTER,除非它的優先級(priority)更高。當MASTER不可用時(BACKUP收不到通告信息), 多臺BACKUP中優先級最高的這臺會被搶佔爲MASTER。這種搶佔是很是快速的(<1s),以保證服務的連續性。因爲安全性考慮,VRRP包使用了加密協議進行加密。

4.VRRP 工做流程

(1).初始化:    
路由器啓動時,若是路由器的優先級是255(最高優先級,路由器擁有路由器地址),要發送VRRP通告信息,併發送廣播ARP信息通告路由器IP地址對應的MAC地址爲路由虛擬MAC,設置通告信息定時器準備定時發送VRRP通告信息,轉爲MASTER狀態;不然進入BACKUP狀態,設置定時器檢查定時檢查是否收到MASTER的通告信息。

(2).Master

  • 設置定時通告定時器;

  • 用VRRP虛擬MAC地址響應路由器IP地址的ARP請求;

  • 轉發目的MAC是VRRP虛擬MAC的數據包;

  • 若是是虛擬路由器IP的擁有者,將接受目的地址是虛擬路由器IP的數據包,不然丟棄;

  • 當收到shutdown的事件時刪除定時通告定時器,發送優先權級爲0的通告包,轉初始化狀態;

  • 若是定時通告定時器超時時,發送VRRP通告信息;

  • 收到VRRP通告信息時,若是優先權爲0,發送VRRP通告信息;不然判斷數據的優先級是否高於本機,或相等並且實際IP地址大於本地實際IP,設置定時通告定時器,復位主機超時定時器,轉BACKUP狀態;不然的話,丟棄該通告包;

(3).Backup

  • 設置主機超時定時器;

  • 不能響應針對虛擬路由器IP的ARP請求信息;

  • 丟棄全部目的MAC地址是虛擬路由器MAC地址的數據包;

  • 不接受目的是虛擬路由器IP的全部數據包;

  • 當收到shutdown的事件時刪除主機超時定時器,轉初始化狀態;

  • 主機超時定時器超時的時候,發送VRRP通告信息,廣播ARP地址信息,轉MASTER狀態;

  • 收到VRRP通告信息時,若是優先權爲0,表示進入MASTER選舉;不然判斷數據的優先級是否高於本機,若是高的話認可MASTER有效,復位主機超時定時器;不然的話,丟棄該通告包;

5.ARP查詢處理

       當內部主機經過ARP查詢虛擬路由器IP地址對應的MAC地址時,MASTER路由器回覆的MAC地址爲虛擬的VRRP的MAC地址,而不是實際網卡的 MAC地址,這樣在路由器切換時讓內網機器覺察不到;而在路由器從新啓動時,不能主動發送本機網卡的實際MAC地址。若是虛擬路由器開啓的ARP代理 (proxy_arp)功能,代理的ARP迴應也迴應VRRP虛擬MAC地址;好了VRRP的簡單講解就到這裏,咱們下來說解一下Keepalived的案例。

3、環境準備

1.操做系統

  • CentOS 6.4 X86_64

2.軟件版本

  • ipvsadm.x86_64 0:1.25-10.el6                  

  • keepalived.x86_64 0:1.2.7-3.el6      

  • httpd-2.2.15-29.el6.centos.x86_64

3.實驗拓撲

k00

4.時間同步

node1:

1
[root@node1 ~] # ntpdate 202.120.2.101

node2:

1
[root@node2 ~] # ntpdate 202.120.2.101

master:

1
[root@master ~] # ntpdate 202.120.2.101

slave:

1
[root@slave ~] # ntpdate 202.120.2.101

5.主機名互相解析

node1:

1
2
3
4
5
[root@node1 ~] # cat /etc/hosts 
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4  
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6  
192.168.18.201    node1. test .com    node1  
192.168.18.202    node2. test .com    node2

node2:

1
2
3
4
5
[root@node2 ~] # cat /etc/hosts 
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4  
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6  
192.168.18.201    node1. test .com    node1  
192.168.18.202    node2. test .com    node2

6.安裝yum源

node1:

1
2
[root@node1 ~] # rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm
[root@node1 ~] # rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

node2:

1
2
[root@node2 ~] # rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm
[root@node2 ~] # rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

master:

1
2
[root@master ~] # rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm
[root@master ~] # rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

slave:

1
2
[root@slave ~] # rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm
[root@slave ~] # rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

4、LVS+Keepalived 實現高可用的前端負載均衡器

node1:

1.安裝httpd

1
[root@node1 ~] # yum install -y httpd

2.配置httpd

1
2
[root@node1 ~] # vim /var/www/html/index.html
<h1>RS1. test .com< /h1 >

3.啓動httpd

1
[root@node1 ~] # service httpd start

4.測試

k1

5.設置開機自啓動

1
2
3
[root@node1 ~] # chkconfig httpd on  
[root@node1 ~] # chkconfig httpd --list   
httpd              0:關閉    1:關閉    2:啓用    3:啓用    4:啓用    5:啓用    6:關閉

6.配置node1

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
[root@node1 ~] # mkdir src  
[root@node1 ~] # cd src/   
[root@node1 src] # vim realserver.sh
#!/bin/bash  
#   
# Script to start LVS DR real server.   
# description: LVS DR real server   
#   
/etc/rc .d /init .d /functions
VIP=192.168.18.200 #修改你的VIP  
host=` /bin/hostname `
case "$1" in  
start)   
        # Start LVS-DR real server on this machine.   
         /sbin/ifconfig lo down   
         /sbin/ifconfig lo up   
         echo 1 > /proc/sys/net/ipv4/conf/lo/arp_ignore   
         echo 2 > /proc/sys/net/ipv4/conf/lo/arp_announce   
         echo 1 > /proc/sys/net/ipv4/conf/all/arp_ignore   
         echo 2 > /proc/sys/net/ipv4/conf/all/arp_announce
         /sbin/ifconfig lo:0 $VIP broadcast $VIP netmask 255.255.255.255 up  
         /sbin/route add -host $VIP dev lo:0
;;  
stop)
         # Stop LVS-DR real server loopback device(s).  
         /sbin/ifconfig lo:0 down   
         echo 0 > /proc/sys/net/ipv4/conf/lo/arp_ignore   
         echo 0 > /proc/sys/net/ipv4/conf/lo/arp_announce   
         echo 0 > /proc/sys/net/ipv4/conf/all/arp_ignore   
         echo 0 > /proc/sys/net/ipv4/conf/all/arp_announce
;;  
status)
         # Status of LVS-DR real server.  
         islothere=` /sbin/ifconfig lo:0 | grep $VIP`   
         isrothere=` netstat -rn | grep "lo:0" | grep $VIP`   
         if [ ! "$islothere" -o ! "isrothere" ]; then   
             # Either the route or the lo:0 device   
             # not found.   
             echo "LVS-DR real server Stopped."   
         else   
             echo "LVS-DR real server Running."   
         fi   
;;   
*)   
             # Invalid entry.   
             echo "$0: Usage: $0 {start|status|stop}"   
             exit 1   
;;   
esac   
[root@node1 src] # chmod +x realserver.sh  
[root@node1 src] # ll   
總用量 4   
-rwxr-xr-x 1 root root 1485 8月  22 10:18 realserver.sh
[root@node1 src] # ./realserver.sh start

7.查看配置

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
[root@node1 src] # ifconfig  
eth0      Link encap:Ethernet  HWaddr 00:0C:29:11:92:E4
           inet addr:192.168.18.201  Bcast:192.168.18.255  Mask:255.255.255.0   
           inet6 addr: fe80::20c:29ff:fe11:92e4 /64 Scope:Link   
           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1   
           RX packets:115061 errors:0 dropped:0 overruns:0 frame:0   
           TX packets:14979 errors:0 dropped:0 overruns:0 carrier:0   
           collisions:0 txqueuelen:1000   
           RX bytes:43448483 (41.4 MiB)  TX bytes:1224926 (1.1 MiB)
lo        Link encap:Local Loopback
           inet addr:127.0.0.1  Mask:255.0.0.0   
           inet6 addr: ::1 /128 Scope:Host   
           UP LOOPBACK RUNNING  MTU:16436  Metric:1   
           RX packets:2 errors:0 dropped:0 overruns:0 frame:0   
           TX packets:2 errors:0 dropped:0 overruns:0 carrier:0   
           collisions:0 txqueuelen:0   
           RX bytes:168 (168.0 b)  TX bytes:168 (168.0 b)
lo:0      Link encap:Local Loopback
           inet addr:192.168.18.200  Mask:255.255.255.255   
           UP LOOPBACK RUNNING  MTU:16436  Metric:1
[root@node1 src] # route -n  
Kernel IP routing table   
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface   
192.168.18.200  0.0.0.0         255.255.255.255 UH    0      0        0 lo   
192.168.18.0    0.0.0.0         255.255.255.0   U     0      0        0 eth0   
169.254.0.0     0.0.0.0         255.255.0.0     U     1002   0        0 eth0   
0.0.0.0         192.168.18.254  0.0.0.0         UG    0      0        0 eth0
[root@node1 src] # cat /proc/sys/net/ipv4/conf/lo/arp_ignore  
1   
[root@node1 src] # cat /proc/sys/net/ipv4/conf/lo/arp_announce   
2   
[root@node1 src] # cat /proc/sys/net/ipv4/conf/all/arp_ignore   
1   
[root@node1 src] # cat /proc/sys/net/ipv4/conf/all/arp_announce
2

好了,node1到這裏基本配置完成,下面咱們來配置node2。

node2:

1.安裝httpd

1
[root@node2 ~] # yum install -y httpd

2.配置httpd

1
2
[root@node2 ~] # vim /var/www/html/index.html
<h1>RS2. test .com< /h1 >

3.啓動httpd

1
[root@node2 ~] # service httpd start

4.測試

k2

5.設置開機自啓動

1
2
3
[root@node2 ~] # chkconfig httpd on  
[root@node2 ~] # chkconfig httpd --list   
httpd              0:關閉    1:關閉    2:啓用    3:啓用    4:啓用    5:啓用    6:關閉

6.配置node2

 

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
[root@node2 ~] # mkdir src  
[root@node2 ~] # cd src/   
[root@node2 src] # vim realserver.sh
#!/bin/bash  
#   
# Script to start LVS DR real server.   
# description: LVS DR real server   
#   
/etc/rc .d /init .d /functions
VIP=192.168.18.200  
host=` /bin/hostname `
case "$1" in  
start)   
        # Start LVS-DR real server on this machine.   
         /sbin/ifconfig lo down   
         /sbin/ifconfig lo up   
         echo 1 > /proc/sys/net/ipv4/conf/lo/arp_ignore   
         echo 2 > /proc/sys/net/ipv4/conf/lo/arp_announce   
         echo 1 > /proc/sys/net/ipv4/conf/all/arp_ignore   
         echo 2 > /proc/sys/net/ipv4/conf/all/arp_announce
         /sbin/ifconfig lo:0 $VIP broadcast $VIP netmask 255.255.255.255 up  
         /sbin/route add -host $VIP dev lo:0
;;  
stop)
         # Stop LVS-DR real server loopback device(s).  
         /sbin/ifconfig lo:0 down   
         echo 0 > /proc/sys/net/ipv4/conf/lo/arp_ignore   
         echo 0 > /proc/sys/net/ipv4/conf/lo/arp_announce   
         echo 0 > /proc/sys/net/ipv4/conf/all/arp_ignore   
         echo 0 > /proc/sys/net/ipv4/conf/all/arp_announce
;;  
status)
         # Status of LVS-DR real server.  
         islothere=` /sbin/ifconfig lo:0 | grep $VIP`   
         isrothere=` netstat -rn | grep "lo:0" | grep $VIP`   
         if [ ! "$islothere" -o ! "isrothere" ]; then   
             # Either the route or the lo:0 device   
             # not found.   
             echo "LVS-DR real server Stopped."   
         else   
             echo "LVS-DR real server Running."   
         fi   
;;   
*)   
             # Invalid entry.   
             echo "$0: Usage: $0 {start|status|stop}"   
             exit 1   
;;   
esac
[root@node2 src] # chmod +x realserver.sh  
[root@node2 src] # ./realserver.sh start

7.查看配置

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
[root@node2 src] # ifconfig  
eth0      Link encap:Ethernet  HWaddr 00:0C:29:B8:DF:6A
           inet addr:192.168.18.202  Bcast:192.168.18.255  Mask:255.255.255.0   
           inet6 addr: fe80::20c:29ff:feb8:df6a /64 Scope:Link   
           UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1   
           RX packets:110545 errors:0 dropped:0 overruns:0 frame:0   
           TX packets:10461 errors:0 dropped:0 overruns:0 carrier:0   
           collisions:0 txqueuelen:1000   
           RX bytes:32853789 (31.3 MiB)  TX bytes:889109 (868.2 KiB)
lo        Link encap:Local Loopback
           inet addr:127.0.0.1  Mask:255.0.0.0   
           inet6 addr: ::1 /128 Scope:Host   
           UP LOOPBACK RUNNING  MTU:16436  Metric:1   
           RX packets:6 errors:0 dropped:0 overruns:0 frame:0   
           TX packets:6 errors:0 dropped:0 overruns:0 carrier:0   
           collisions:0 txqueuelen:0   
           RX bytes:616 (616.0 b)  TX bytes:616 (616.0 b)
lo:0      Link encap:Local Loopback
           inet addr:192.168.18.200  Mask:255.255.255.255   
           UP LOOPBACK RUNNING  MTU:16436  Metric:1
[root@node2 src] # route -n  
Kernel IP routing table   
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface   
192.168.18.200  0.0.0.0         255.255.255.255 UH    0      0        0 lo   
192.168.18.0    0.0.0.0         255.255.255.0   U     0      0        0 eth0   
169.254.0.0     0.0.0.0         255.255.0.0     U     1002   0        0 eth0   
0.0.0.0         192.168.18.254  0.0.0.0         UG    0      0        0 eth0
[root@node2 src] # cat /proc/sys/net/ipv4/conf/lo/arp_ignore  
1   
[root@node2 src] # cat /proc/sys/net/ipv4/conf/lo/arp_announce   
2   
[root@node2 src] # cat /proc/sys/net/ipv4/conf/all/arp_ignore   
1   
[root@node2 src] # cat /proc/sys/net/ipv4/conf/all/arp_announce   
2

好了,到這裏node2也基本配置完成。下面咱們來配置master與slave。

masterg與slave:

1.安裝keepalived與ipvsadm

1
2
[root@master ~] # yum install -y keepalived ipvsadm
[root@slave ~] # yum install -y keepalived ipvsadm

2.修改配置文件

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
[root@master ~] # cat /etc/keepalived/keepalived.conf  
! Configuration File for keepalived
global_defs {  
    notification_email {   
15251076067@163.com #配置管理員郵箱   
    }   
    notification_email_from root #配置發件人   
    smtp_server 127.0.0.1 #配置郵件服務器   
    smtp_connect_timeout 30   
    router_id LVS_DEVEL   
}
vrrp_instance VI_1 {  
     state MASTER #配置模式   
     interface eth0   
     virtual_router_id 51   
     priority 101 #配置優先級   
     advert_int 1   
     authentication {   
         auth_type PASS   
         auth_pass 1111   
     }   
     virtual_ipaddress {   
         192.168.18.200 #配置虛擬IP地址   
     }   
}
virtual_server 192.168.18.200 80 {  
     delay_loop 6   
     lb_algo rr   
     lb_kind DR   
     nat_mask 255.255.255.0   
     #persistence_timeout 50   
     protocol TCP
     real_server 192.168.18.201 80 { #配置realaserver  
         weight 1   
         HTTP_GET { #監控配置   
             url {   
               path /   
           status_code 200   
             }   
             connect_timeout 2   
             nb_get_retry 3   
             delay_before_retry 1   
         }   
     }   
     real_server 192.168.18.202 80 {   
         weight 1   
         HTTP_GET {   
             url {   
               path /   
               status_code 200   
             }   
             connect_timeout 2   
             nb_get_retry 3   
             delay_before_retry 1   
         }   
     }   
}

3.將配置文件同步到slave

1
[root@master ~] # scp /etc/keepalived/keepalived.conf root@192.168.18.207:/etc/keepalived/

4.簡單修改一下slave配置文件

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
[root@slave keepalived] # cat keepalived.conf  
! Configuration File for keepalived
global_defs {  
    notification_email {   
      15251076067@163.com   
    }   
    notification_email_from root   
    smtp_server 127.0.0.1   
    smtp_connect_timeout 30   
    router_id LVS_DEVEL   
}
vrrp_instance VI_1 {  
     state BACKUP #修改成BACKUP   
     interface eth0   
     virtual_router_id 51   
     priority 100 #修改優先級   
     advert_int 1   
     authentication {   
         auth_type PASS   
         auth_pass 1111   
     }   
     virtual_ipaddress {   
         192.168.18.200   
     }   
}
virtual_server 192.168.18.200 80 {  
     delay_loop 6   
     lb_algo rr   
     lb_kind DR   
     nat_mask 255.255.255.0   
     #persistence_timeout 50   
     protocol TCP
     real_server 192.168.18.201 80 {  
         weight 1   
         HTTP_GET {   
             url {   
               path /   
           status_code 200   
             }   
             connect_timeout 2   
             nb_get_retry 3   
             delay_before_retry 1   
         }   
     }   
     real_server 192.168.18.202 80 {   
         weight 1   
         HTTP_GET {   
             url {   
               path /   
               status_code 200   
             }   
             connect_timeout 2   
             nb_get_retry 3   
             delay_before_retry 1   
         }   
     }   
}

5.啓動master與slave的keepalived服務

1
2
3
4
[root@master ~] # service keepalived start  
正在啓動 keepalived:                                      [肯定]
[root@slave ~] # service keepalived start  
正在啓動 keepalived:                                      [肯定]

6.查看一下LVS狀態

1
2
3
4
5
6
7
[root@master ~] # ipvsadm -L -n  
IP Virtual Server version 1.2.1 (size=4096)   
Prot LocalAddress:Port Scheduler Flags   
   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn   
TCP  192.168.18.200:80 rr   
   -> 192.168.18.201:80            Route   1      0          0       
   -> 192.168.18.202:80            Route   1      0          0

7.測試

k3

k4

8.模擬故障

(1).中止一下node1

1
2
[root@node1 src] # service httpd stop  
中止 httpd:                                               [肯定]

(2).查看一下的lvs

1
2
3
4
5
6
[root@master ~] # ipvsadm -L -n  
IP Virtual Server version 1.2.1 (size=4096)   
Prot LocalAddress:Port Scheduler Flags   
   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn   
TCP  192.168.18.200:80 rr   
   -> 192.168.18.202:80            Route   1      0          0

(3).測試一下

k4

(4).查看一下郵件

k5

(5).從新啓動一下node1

1
2
[root@node1 src] # service httpd start  
正在啓動 httpd:                                           [肯定]

(6).再查看一下lvs狀態

1
2
3
4
5
6
7
[root@master ~] # ipvsadm -L -n  
IP Virtual Server version 1.2.1 (size=4096)   
Prot LocalAddress:Port Scheduler Flags   
   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn   
TCP  192.168.18.200:80 rr   
   -> 192.168.18.201:80            Route   1      0          0       
   -> 192.168.18.202:80            Route   1      0          0

(7).再查看一下郵件

k6

(8).關閉master上keepalived

1
2
3
4
5
6
[root@master ~] # service keepalived stop  
中止 keepalived:                                          [肯定]   
[root@master ~] # ipvsadm -L -n   
IP Virtual Server version 1.2.1 (size=4096)   
Prot LocalAddress:Port Scheduler Flags   
   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn

(9).查看一下slave狀態

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
[root@slave ~] # ip addr show  
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN   
     link /loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00   
     inet 127.0.0.1 /8 scope host lo   
     inet6 ::1 /128 scope host   
        valid_lft forever preferred_lft forever   
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000   
     link /ether 00:0c:29:f9:e6:26 brd ff:ff:ff:ff:ff:ff   
     inet 192.168.18.207 /24 brd 192.168.18.255 scope global eth0   
     inet 192.168.18.200 /32 scope global eth0   
     inet6 fe80::20c:29ff:fef9:e626 /64 scope link   
        valid_lft forever preferred_lft forever   
[root@slave ~] # ipvsadm -L -n   
IP Virtual Server version 1.2.1 (size=4096)   
Prot LocalAddress:Port Scheduler Flags   
   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn   
TCP  192.168.18.200:80 rr   
   -> 192.168.18.201:80            Route   1      0          0       
   -> 192.168.18.202:80            Route   1      0          0

(10).再次測試一下

k3

注,你們能夠看到,通過上面的演示咱們如今LVS的高可用即前端負載均衡的高可用,同時實現對後端realserver監控,也實現後端realserver宕機時會給管理員發送郵件。但還有幾個問題咱們尚未解決,問題以下:

  • 全部realserver都down機,怎麼處理?是否是用戶就無法打開,仍是提供一下維護頁面。

  • 怎麼完成維護模式keepalived切換?

  • 如何在keepalived故障時,發送警告郵件給指定的管理員?

9.全部realserver都down機,怎麼處理?

問題:在集羣中若是全部real server所有宕機了,客戶端訪問時就會出現錯誤頁面,這樣是很不友好的,咱們得提供一個維護頁面來提醒用戶,服務器正在維護,什麼時間能夠訪問等,下面咱們就來解決一下這個問題。解決方案有兩種,一種是提供一臺備用的real server當全部的服務器宕機時,提供維護頁面,但這樣作有點浪費服務器。另外一種就是在負載均衡器上提供維護頁面,這樣是比較靠譜的,也比較經常使用。下面咱們就來具體操做一下。

(1).master與slave安裝上httpd

1
2
[root@master ~] # yum install -y httpd
[root@slave ~] # yum install -y httpd

(2).配置維護頁面

1
2
3
4
[root@master ~] # vim /var/www/html/index.html 
Website is currently under maintenance, please come back later!
  [root@slave ~] # vim /var/www/html/index.html
Website is currently under maintenance, please come back later!

(3).啓動httpd服務並測試

1
2
3
4
[root@master ~] # service httpd start 
正在啓動 httpd:                                           [肯定]
[root@slave ~] # service httpd start 
正在啓動 httpd:                                           [肯定]

k10

k11

(4).修改配置文件

master:

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
[root@master ~] # cat /etc/keepalived/keepalived.conf  
! Configuration File for keepalived
global_defs { 
    notification_email {  
      15251076067@163.com  
    }  
    notification_email_from root  
    smtp_server 127.0.0.1  
    smtp_connect_timeout 30  
    router_id LVS_DEVEL  
}
vrrp_instance VI_1 { 
     state MASTER  
     interface eth0  
     virtual_router_id 51  
     priority 101  
     advert_int 1  
     authentication {  
         auth_type PASS  
         auth_pass 1111  
     }  
     virtual_ipaddress {  
         192.168.18.200  
     }  
}
virtual_server 192.168.18.200 80 { 
     delay_loop 6  
     lb_algo rr  
     lb_kind DR  
     nat_mask 255.255.255.0  
     protocol TCP
     real_server 192.168.18.201 80 { 
         weight 1  
         HTTP_GET {  
             url {  
               path /  
           status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     real_server 192.168.18.202 80 {  
         weight 1  
         HTTP_GET {  
             url {  
               path /  
               status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
    sorry_server 127.0.0.1 80   #增長一行sorry_server   
}

slave:

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
[root@slave ~] # cat /etc/keepalived/keepalived.conf  
! Configuration File for keepalived
global_defs { 
    notification_email {  
      15251076067@163.com  
    }  
    notification_email_from root  
    smtp_server 127.0.0.1  
    smtp_connect_timeout 30  
    router_id LVS_DEVEL  
}
vrrp_instance VI_1 { 
     state BACKUP  
     interface eth0  
     virtual_router_id 51  
     priority 100  
     advert_int 1  
     authentication {  
         auth_type PASS  
         auth_pass 1111  
     }  
     virtual_ipaddress {  
         192.168.18.200  
     }  
}
virtual_server 192.168.18.200 80 { 
     delay_loop 6  
     lb_algo rr  
     lb_kind DR  
     nat_mask 255.255.255.0  
     protocol TCP
     real_server 192.168.18.201 80 { 
         weight 1  
         HTTP_GET {  
             url {  
               path /  
           status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     real_server 192.168.18.202 80 {  
         weight 1  
         HTTP_GET {  
             url {  
               path /  
               status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     sorry_server 127.0.0.1 80 #增長一行sorry_server   
}

(5).關閉全部的real server並從新啓動一下master與slave的keepalived

1
2
3
4
5
6
7
8
9
10
[root@node1 ~] # service httpd stop 
中止 httpd:                                               [肯定]
[root@node2 ~] # service httpd stop 
中止 httpd:                                               [肯定]
[root@master ~] # service keepalived restart 
中止 keepalived:                                          [肯定]  
正在啓動 keepalived:                                      [肯定]
[root@slave ~] # service keepalived restart 
中止 keepalived:                                          [肯定]  
正在啓動 keepalived:                                      [肯定]

(6).查看一下lvs

1
2
3
4
5
6
[root@master ~] # ipvsadm -L -n 
IP Virtual Server version 1.2.1 (size=4096)  
Prot LocalAddress:Port Scheduler Flags  
   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn  
TCP  192.168.18.200:80 rr  
   -> 127.0.0.1:80                 Local   1      0          0

(7).測試

k12

注,sorry_server測試成功,下面咱們繼續。

10.怎麼完成維護模式keepalived切換?

問題:咱們通常進行主從切換測試時都是關閉keepalived或關閉網卡接口,有沒有一種方法能實如今不關閉keepalived下或網卡接口來實現維護呢?方法確定是有的,在keepalived新版本中,支持腳本vrrp_srcipt,具體如何使用你們能夠man keepalived.conf查看。下面咱們來演示一下具體怎麼實現。

(1).定義腳本

1
2
3
4
5
6
7
vrrp_srcipt chk_schedown { 
    script "[ -e /etc/keepalived/down ] && exit 1 || exit 0"  
    interval 1 #監控間隔  
    weight -5 #減少優先級  
    fall 2 #監控失敗次數  
    rise 1 #監控成功次數  
}

(2).執行腳本

1
2
3
track_script {
    chk_schedown #執行chk_schedown腳本
}

(3).修改配置文件

master:

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
[root@master ~] # cat /etc/keepalived/keepalived.conf  
! Configuration File for keepalived
global_defs { 
    notification_email {  
      15251076067@163.com  
    }  
    notification_email_from root  
    smtp_server 127.0.0.1  
    smtp_connect_timeout 30  
    router_id LVS_DEVEL  
}
vrrp_script chk_schedown { #定義vrrp執行腳本  
    script "[ -e /etc/keepalived/down ] && exit 1 || exit 0" #查看是否有down文件,有就進入維護模式  
    interval 1 #監控間隔時間  
    weight -5 #下降優先級  
    fall 2 #失敗次數  
    rise 1 #成功數次  
}
vrrp_instance VI_1 { 
     state MASTER  
     interface eth0  
     virtual_router_id 51  
     priority 101  
     advert_int 1  
     authentication {  
         auth_type PASS  
         auth_pass 1111  
     }  
     virtual_ipaddress {  
         192.168.18.200  
     }  
     track_script { #執行腳本  
         chk_schedown  
     }  
}
virtual_server 192.168.18.200 80 { 
     delay_loop 6  
     lb_algo rr  
     lb_kind DR  
     nat_mask 255.255.255.0  
     protocol TCP
     real_server 192.168.18.201 80 { 
         weight 1  
         HTTP_GET {  
             url {  
               path /  
           status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     real_server 192.168.18.202 80 {  
         weight 1  
         HTTP_GET {  
             url {  
               path /  
               status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
    sorry_server 127.0.0.1 80  
}

slave:

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
[root@slave ~] # cat /etc/keepalived/keepalived.conf  
! Configuration File for keepalived
global_defs { 
    notification_email {  
      15251076067@163.com  
    }  
    notification_email_from root  
    smtp_server 127.0.0.1  
    smtp_connect_timeout 30  
    router_id LVS_DEVEL  
}
vrrp_script chk_schedown { 
    script "[ -e /etc/keepalived/down ] && exit 1 || exit 0"  
    interval 1  
    weight -5  
    fall 2  
    rise 1  
}
vrrp_instance VI_1 { 
     state BACKUP  
     interface eth0  
     virtual_router_id 51  
     priority 100  
     advert_int 1  
     authentication {  
         auth_type PASS  
         auth_pass 1111  
     }  
     virtual_ipaddress {  
         192.168.18.200  
     }  
     track_script {  
     chk_schedown  
     }
}
virtual_server 192.168.18.200 80 { 
     delay_loop 6  
     lb_algo rr  
     lb_kind DR  
     nat_mask 255.255.255.0  
     protocol TCP
     real_server 192.168.18.201 80 { 
         weight 1  
         HTTP_GET {  
             url {  
               path /  
           status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     real_server 192.168.18.202 80 {  
         weight 1  
         HTTP_GET {  
             url {  
               path /  
               status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     sorry_server 127.0.0.1 80

(4).測試

master:

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
[root@master keepalived] # touch down #新建一下down文件
[root@master keepalived] # ll 
總用量 4  
-rw-r--r-- 1 root root    0 8月  22 13:39 down  
-rw-r--r-- 1 root root 1317 8月  22 13:35 keepalived.conf
[root@master keepalived] # tail -f /var/log/messages #查看一下日誌 
Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) Entering MASTER STATE  
Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) setting protocol VIPs.  
Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.18.200  
Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) Received higher prio advert  
Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) Entering BACKUP STATE  
Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) removing protocol VIPs.  
Aug 22 13:43:52 master Keepalived_healthcheckers[12002]: Netlink reflector reports IP 192.168.18.200 added  
Aug 22 13:43:52 master Keepalived_healthcheckers[12002]: Netlink reflector reports IP 192.168.18.200 removed  
Aug 22 13:43:52 master Keepalived_healthcheckers[12002]: SMTP alert successfully sent.  
Aug 22 13:43:52 master Keepalived_healthcheckers[12002]: SMTP alert successfully sent.  
^C  
[root@master keepalived] # ip add show #查看VIP  
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN   
     link /loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00  
     inet 127.0.0.1 /8 scope host lo  
     inet6 ::1 /128 scope host   
        valid_lft forever preferred_lft forever  
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000  
     link /ether 00:0c:29:4b:a1:85 brd ff:ff:ff:ff:ff:ff  
     inet 192.168.18.208 /24 brd 192.168.18.255 scope global eth0  
     inet6 fe80::20c:29ff:fe4b:a185 /64 scope link   
        valid_lft forever preferred_lft forever

slave:

1
2
3
4
5
6
7
8
9
10
11
12
[root@slave ~] # ip addr show #查看一下VIP已轉移到slave上 
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN   
     link /loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00  
     inet 127.0.0.1 /8 scope host lo  
     inet6 ::1 /128 scope host   
        valid_lft forever preferred_lft forever  
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000  
     link /ether 00:0c:29:f9:e6:26 brd ff:ff:ff:ff:ff:ff  
     inet 192.168.18.207 /24 brd 192.168.18.255 scope global eth0  
     inet 192.168.18.200 /32 scope global eth0  
     inet6 fe80::20c:29ff:fef9:e626 /64 scope link   
        valid_lft forever preferred_lft forever

好了,自寫監測腳本,完成維護模式切換,到這裏就演示成功,下面咱們來解決最後一個問題,就是keepalived主從切換的郵件通告。

11.如何在keepalived故障時(或主備切換時),發送警告郵件給指定的管理員?

(1).keepalived通知腳本進階示例

下面的腳本能夠接受選項,其中

  • -s, --service SERVICE,...:指定服務腳本名稱,當狀態切換時可自動啓動、重啓或關閉此服務;

  • -a, --address VIP: 指定相關虛擬路由器的VIP地址;

  • -m, --mode {mm|mb}:指定虛擬路由的模型,mm表示主主,mb表示主備;它們表示相對於同一種服務而方,其VIP的工做類型;

  • -n, --notify {master|backup|fault}:指定通知的類型,即vrrp角色切換的目標角色;

  • -h, --help:獲取腳本的使用幫助;

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
#!/bin/bash 
# Author: freeloda   
# description: An example of notify script  
# Usage: notify.sh -m|--mode {mm|mb} -s|--service SERVICE1,... -a|--address VIP  -n|--notify {master|backup|falut} -h|--help
contact= '1521076067@163.com'
helpflag=0  
serviceflag=0  
modeflag=0  
addressflag=0  
notifyflag=0
Usage() { 
   echo "Usage: notify.sh [-m|--mode {mm|mb}] [-s|--service SERVICE1,...] <-a|--address VIP>  <-n|--notify {master|backup|falut}>"   
   echo "Usage: notify.sh -h|--help"  
}
ParseOptions() { 
   local I=1;  
   if [ $ # -gt 0 ]; then  
     while [ $I - le $ # ]; do  
       case $1 in  
       -s|--service)  
         [ $ # -lt 2 ] && return 3  
          serviceflag=1  
          services=(` echo $2| awk -F "," '{for(i=1;i<=NF;i++) print $i}' `)  
         shift 2 ;;  
       -h|--help)  
          helpflag=1  
         return 0  
         shift  
         ;;  
       -a|--address)  
         [ $ # -lt 2 ] && return 3  
         addressflag=1  
         vip=$2  
         shift 2  
         ;;  
       -m|--mode)  
         [ $ # -lt 2 ] && return 3  
         mode=$2  
         shift 2  
         ;;  
       -n|--notify)  
         [ $ # -lt 2 ] && return 3  
         notifyflag=1  
         notify=$2  
         shift 2  
         ;;  
       *)  
         echo "Wrong options..."  
         Usage  
         return 7  
         ;;  
        esac  
     done  
     return 0  
   fi  
}
#workspace=$(dirname $0)
RestartService() { 
   if [ ${ #@} -gt 0 ]; then  
     for I in $@; do  
       if [ -x /etc/rc .d /init .d/$I ]; then  
         /etc/rc .d /init .d/$I restart  
       else  
         echo "$I is not a valid service..."  
       fi  
     done  
   fi  
}
StopService() { 
   if [ ${ #@} -gt 0 ]; then  
     for I in $@; do  
       if [ -x /etc/rc .d /init .d/$I ]; then  
         /etc/rc .d /init .d/$I stop  
       else  
         echo "$I is not a valid service..."  
       fi  
     done  
   fi  
}
Notify() {  
     mailsubject= "`hostname` to be $1: $vip floating"  
     mailbody= "`date '+%F %H:%M:%S'`, vrrp transition, `hostname` changed to be $1."  
     echo $mailbody | mail -s "$mailsubject" $contact  
}
# Main Function  
ParseOptions $@  
[ $? - ne 0 ] && Usage && exit 5
[ $helpflag - eq 1 ] && Usage && exit 0
if [ $addressflag - ne 1 -o $notifyflag - ne 1 ]; then 
   Usage  
   exit 2  
fi
mode=${mode:-mb}
case $notify in 
'master' )  
   if [ $serviceflag - eq 1 ]; then  
       RestartService ${services[*]}  
   fi  
   Notify master  
   ;;  
'backup' )  
   if [ $serviceflag - eq 1 ]; then  
     if [ "$mode" == 'mb' ]; then  
       StopService ${services[*]}  
     else  
       RestartService ${services[*]}  
     fi  
   fi  
   Notify backup  
   ;;  
'fault' )  
   Notify fault  
   ;;  
*)  
   Usage  
   exit 4  
   ;;  
esac

(2).在keepalived.conf配置文件中,其調用方法以下所示:

  • notify_master "/etc/keepalived/notify.sh -n master -a 192.168.18.200"  

  • notify_backup "/etc/keepalived/notify.sh -n backup -a 192.168.18.200"  

  • notify_fault "/etc/keepalived/notify.sh -n fault -a 192.168.18.200"  

(3).修改配置文件

master:

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
[root@master keepalived] # cat keepalived.conf  
! Configuration File for keepalived
global_defs { 
    notification_email {  
      15251076067@163.com  
    }  
    notification_email_from root  
    smtp_server 127.0.0.1  
    smtp_connect_timeout 30  
    router_id LVS_DEVEL  
}
vrrp_script chk_schedown { 
    script "[ -e /etc/keepalived/down ] && exit 1 || exit 0"  
    interval 1  
    weight -5  
    fall 2  
    rise 1  
}
vrrp_instance VI_1 { 
     state MASTER  
     interface eth0  
     virtual_router_id 51  
     priority 101  
     advert_int 1  
     authentication {  
         auth_type PASS  
         auth_pass 1111  
     }  
     virtual_ipaddress {  
         192.168.18.200  
     }  
     track_script {  
         chk_schedown  
     }
     #增長如下三行 
     notify_master "/etc/keepalived/notify.sh -n master -a 192.168.18.200"  
     notify_backup "/etc/keepalived/notify.sh -n backup -a 192.168.18.200"
     notify_fault "/etc/keepalived/notify.sh -n fault -a 192.168.18.200"
}
virtual_server 192.168.18.200 80 { 
     delay_loop 6  
     lb_algo rr  
     lb_kind DR  
     nat_mask 255.255.255.0  
     protocol TCP
     real_server 192.168.18.201 80 { 
         weight 1  
         HTTP_GET {  
             url {  
               path /  
           status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     real_server 192.168.18.202 80 {  
         weight 1  
         HTTP_GET {  
             url {  
               path /  
               status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
    sorry_server 127.0.0.1 80  
}

slave:

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
[root@slave keepalived] # cat keepalived.conf  
! Configuration File for keepalived
global_defs { 
    notification_email {  
      15251076067@163.com  
    }  
    notification_email_from root  
    smtp_server 127.0.0.1  
    smtp_connect_timeout 30  
    router_id LVS_DEVEL  
}
vrrp_script chk_schedown { 
    script "[ -e /etc/keepalived/down ] && exit 1 || exit 0"  
    interval 1  
    weight -5  
    fall 2  
    rise 1  
}
vrrp_instance VI_1 { 
     state BACKUP  
     interface eth0  
     virtual_router_id 51  
     priority 100  
     advert_int 1  
     authentication {  
         auth_type PASS  
         auth_pass 1111  
     }  
     virtual_ipaddress {  
         192.168.18.200  
     }  
     track_script {  
     chk_schedown  
     }
     #增長如下三行 
     notify_master "/etc/keepalived/notify.sh -n master -a 192.168.18.200"
     notify_backup "/etc/keepalived/notify.sh -n backup -a 192.168.18.200"  
     notify_fault "/etc/keepalived/notify.sh -n fault -a 192.168.18.200"  
}
virtual_server 192.168.18.200 80 { 
     delay_loop 6  
     lb_algo rr  
     lb_kind DR  
     nat_mask 255.255.255.0  
     protocol TCP
     real_server 192.168.18.201 80 { 
         weight 1  
         HTTP_GET {  
             url {  
               path /  
           status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     real_server 192.168.18.202 80 {  
         weight 1  
         HTTP_GET {  
             url {  
               path /  
               status_code 200  
             }  
             connect_timeout 2  
             nb_get_retry 3  
             delay_before_retry 1  
         }  
     }  
     sorry_server 127.0.0.1 80  
}

(4).增長腳本

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
[root@slave keepalived] # pwd 
/etc/keepalived  
[root@slave keepalived] # vim notify.sh
[root@slave keepalived] # cat notify.sh  
#!/bin/bash  
# Author: freeloda   
# description: An example of notify script  
# Usage: notify.sh -m|--mode {mm|mb} -s|--service SERVICE1,... -a|--address VIP  -n|--notify {master|backup|falut} -h|--help
contact= '15251076067@163.com' 
helpflag=0  
serviceflag=0  
modeflag=0  
addressflag=0  
notifyflag=0
Usage() { 
   echo "Usage: notify.sh [-m|--mode {mm|mb}] [-s|--service SERVICE1,...] <-a|--address VIP>  <-n|--notify {master|backup|falut}>"   
   echo "Usage: notify.sh -h|--help"  
}
ParseOptions() { 
   local I=1;  
   if [ $ # -gt 0 ]; then  
     while [ $I - le $ # ]; do  
       case $1 in  
       -s|--service)  
         [ $ # -lt 2 ] && return 3  
          serviceflag=1  
          services=(` echo $2| awk -F "," '{for(i=1;i<=NF;i++) print $i}' `)  
         shift 2 ;;  
       -h|--help)  
          helpflag=1  
         return 0  
         shift  
         ;;  
       -a|--address)  
         [ $ # -lt 2 ] && return 3  
         addressflag=1  
         vip=$2  
         shift 2  
         ;;  
       -m|--mode)  
         [ $ # -lt 2 ] && return 3  
         mode=$2  
         shift 2  
         ;;  
       -n|--notify)  
         [ $ # -lt 2 ] && return 3  
         notifyflag=1  
         notify=$2  
         shift 2  
         ;;  
       *)  
         echo "Wrong options..."  
         Usage  
         return 7  
         ;;  
        esac  
     done  
     return 0  
   fi  
}
#workspace=$(dirname $0)
RestartService() { 
   if [ ${ #@} -gt 0 ]; then  
     for I in $@; do  
       if [ -x /etc/rc .d /init .d/$I ]; then  
         /etc/rc .d /init .d/$I restart  
       else  
         echo "$I is not a valid service..."  
       fi  
     done  
   fi  
}
StopService() { 
   if [ ${ #@} -gt 0 ]; then  
     for I in $@; do  
       if [ -x /etc/rc .d /init .d/$I ]; then  
         /etc/rc .d /init .d/$I stop  
       else  
         echo "$I is not a valid service..."  
       fi  
     done  
   fi  
}
Notify() {  
     mailsubject= "`hostname` to be $1: $vip floating"  
     mailbody= "`date '+%F %H:%M:%S'`, vrrp transition, `hostname` changed to be $1."  
     echo $mailbody | mail -s "$mailsubject" $contact  
}
# Main Function  
ParseOptions $@  
[ $? - ne 0 ] && Usage && exit 5
[ $helpflag - eq 1 ] && Usage && exit 0
if [ $addressflag - ne 1 -o $notifyflag - ne 1 ]; then 
   Usage  
   exit 2  
fi
mode=${mode:-mb}
case $notify in 
'master' )  
   if [ $serviceflag - eq 1 ]; then  
       RestartService ${services[*]}  
   fi  
   Notify master  
   ;;  
'backup' )  
   if [ $serviceflag - eq 1 ]; then  
     if [ "$mode" == 'mb' ]; then  
       StopService ${services[*]}  
     else  
       RestartService ${services[*]}  
     fi  
   fi  
   Notify backup  
   ;;  
'fault' )  
   Notify fault  
   ;;  
*)  
   Usage  
   exit 4  
   ;;  
esac

(5).給腳本增長執行權限

1
[root@slave keepalived] # chmod +x notify.sh

(6).將master上腳本複製到slave上

1
[root@slave keepalived] # scp -p notify.sh root@192.168.18.207:/etc/keepalived/

(7).測試一下腳本

1
2
3
4
5
6
7
[root@slave keepalived] # ./notify.sh -h 
Usage: notify.sh [-m|--mode {mm|mb}] [-s|--service SERVICE1,...] <-a|--address VIP>  <-n|--notify {master|backup|falut}>  
Usage: notify.sh -h|--help  
[root@slave keepalived] # ./notify.sh --help  
Usage: notify.sh [-m|--mode {mm|mb}] [-s|--service SERVICE1,...] <-a|--address VIP>  <-n|--notify {master|backup|falut}>  
Usage: notify.sh -h|--help  
[root@slave keepalived] # ./notify.sh -m mb -a 1.1.1.1 -n master

(8).查看一下郵件

k13

注,你們能夠看到成功收到郵件,測試成功。在模擬故障時先重啓一下keepalived服務。

(9).模擬故障

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
[root@master keepalived] # ip addr show #查看一下VIP 
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN   
     link /loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00  
     inet 127.0.0.1 /8 scope host lo  
     inet6 ::1 /128 scope host   
        valid_lft forever preferred_lft forever  
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000  
     link /ether 00:0c:29:4b:a1:85 brd ff:ff:ff:ff:ff:ff  
     inet 192.168.18.208 /24 brd 192.168.18.255 scope global eth0  
     inet 192.168.18.200 /32 scope global eth0  
     inet6 fe80::20c:29ff:fe4b:a185 /64 scope link   
        valid_lft forever preferred_lft forever  
[root@master keepalived] # touch down #進入維護模式  
[root@master keepalived] # ll  
總用量 8  
-rw-r--r-- 1 root root    0 8月  22 14:39 down  
-rw-r--r-- 1 root root 1543 8月  22 14:04 keepalived.conf  
-rwxr-xr-x 1 root root 2516 8月  22 14:15 notify.sh  
[root@master keepalived] # ip addr show #再次查看VIP  
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN   
     link /loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00  
     inet 127.0.0.1 /8 scope host lo  
     inet6 ::1 /128 scope host   
        valid_lft forever preferred_lft forever  
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000  
     link /ether 00:0c:29:4b:a1:85 brd ff:ff:ff:ff:ff:ff  
     inet 192.168.18.208 /24 brd 192.168.18.255 scope global eth0  
     inet6 fe80::20c:29ff:fe4b:a185 /64 scope link   
        valid_lft forever preferred_lft forever
[root@slave keepalived] # ip addr show #你們能夠看到VIP成功移動到slave上 
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN   
     link /loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00  
     inet 127.0.0.1 /8 scope host lo  
     inet6 ::1 /128 scope host   
        valid_lft forever preferred_lft forever  
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000  
     link /ether 00:0c:29:f9:e6:26 brd ff:ff:ff:ff:ff:ff  
     inet 192.168.18.207 /24 brd 192.168.18.255 scope global eth0  
     inet 192.168.18.200 /32 scope global eth0  
     inet6 fe80::20c:29ff:fef9:e626 /64 scope link   
        valid_lft forever preferred_lft forever

(10).查看一下郵件

k14

注,你們能夠看到,主備切換時,會發送郵件報警,好了到這裏全部演示所有完成。但願你們有所收穫^_^……

 

本文出自 「Share your knowledge …」 博客,請務必保留此出處http://freeloda.blog.51cto.com/2033581/1280962

相關文章
相關標籤/搜索