Keeplived簡介
keepalived觀其名可知，保持存活，在網絡裏面就是保持在線了，也就是所謂的高可用或熱備，用來防止單點故障(單點故障是指一旦某一點出現故障就會導致整個系統架構的不可用)的發生。
Keepalived原理
keepalived也是模塊化設計，不同模塊複雜不同的功能
core：是keepalived的核心，複雜主進程的啓動和維護，全局配置文件的加載解析等
check：負責healthchecker(健康檢查)，包括了各種健康檢查方式，以及對應的配置的解析包括LVS的配置解析
vrrp：VRRPD子進程，VRRPD子進程就是來實現VRRP協議的
libipfwc：iptables(ipchains)庫，配置LVS會用到
libipvs*：配置LVS會用到
注意，keepalived和LVS完全是兩碼事，只不過他們各負其責相互配合而已
keepalived啓動後會有三個進程
父進程：內存管理，子進程管理等等
子進程：VRRP子進程
子進程：healthchecker子進程
有圖可知，兩個子進程都被系統WatchDog看管，兩個子進程各自複雜自己的事，healthchecker子進程複雜檢查各自服務器的健康程度，例如HTTP，LVS等等，如果healthchecker子進程檢查到MASTER上服務不可用了，就會通知本機上的兄弟VRRP子進程，讓他刪除通告，並且去掉虛擬IP，轉換爲BACKUP狀態
實驗：
主機1 172.16.31.10
主機2 172.16.31.30
1， 下載rpm包至本地，然後安裝，此次實驗用的是keepalived-1.2.13-1.el6.x86_64.rpm主機1,2都安裝
#yum install -y keepalived-1.2.13-1.el6.x86_64.rpm
2， 編輯配置文件在/etc/keepalived/keepalived.conf，一般在配置前最好先把配置文件備份一下
#cp keepalived.conf keepalived.conf.bak
!Configuration File for keepalived //歎號開頭的行爲註釋行
global_defs{ //全局配置
notification_email {
[email protected]
[email protected]
[email protected]
}
notification_email_from [email protected]
smtp_server 192.168.200.1
smtp_connect_timeout 30
router_id LVS_DEVEL
}
vrrp_instanceVI_1 { // 配置實例1
state MASTER //狀態爲主
interface eth0 //在哪個接口上配置
virtual_router_id 51 // 虛擬router_id 號爲51
priority 100 // 優先級
advert_int 1
authentication {
auth_type PASS // 基於密碼的認證
auth_pass1111 // 認證密碼
}
virtual_ipaddress { // 虛擬地址IP組
172.16.31.22
}
}3， 配置備份節點etc/keepalived/keepalived.conf
vrrp_instanceVI_1 {
state BACKUP // 狀態變爲備的，BACKUP
interface eth0
virtual_router_id 101
priority 99 // 優先級比主的小就行
advert_int 1
authentication {
auth_type PASS
auth_pass 123456 // 與主的保持一致
}
virtual_ipaddress {
172.16.31.22
}
}4， 啓動keepalived服務在主機1接點上
# service keepalived start
Starting keepalived: [ OK ]
在主機1上# ip addr show
(1): lo: <LOOPBACK,UP,LOWER_UP> mtu16436 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 qlen1000
link/ether 00:0c:29:a7:90:c5 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.10/16 brd 172.16.255.255 scope global eth0
inet 172.16.31.22/32 scope global eth0 虛擬IP配置上了
inet6 fe80::20c:29ff:fea7:90c5/64 scope link
valid_lft forever preferred_lft forever
5， 然後再主機1上停止一下服務，查看日誌文件
# service keepalived stop
Oct 10 17:00:49 localhost Keepalived[2791]: Stopping Keepalivedv1.2.13 (09/14,2014)
Oct 10 17:00:49 localhost Keepalived_vrrp[2793]:VRRP_Instance(VI_1) removing protocol VIPs.
主機1上# ip addrshow
(1): lo:<LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd00:00:00:00:00:00
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 qlen1000
link/ether 00:0c:29:a7:90:c5 brdff:ff:ff:ff:ff:ff
inet 172.16.31.10/16 brd 172.16.255.255scope global eth0
inet6 fe80::20c:29ff:fea7:90c5/64 scopelink
valid_lftforever preferred_lft forever
此刻172.16.31.22消失了
6， 此刻在啓動2節點keepalived服務，並查看日誌文件
# service keepalived start
Sep 25 08:15:21 localhost Keepalived_vrrp[3954]: VRRP_Instance(VI_1)Transition to MASTER STATE
Sep 25 08:15:22 localhost Keepalived_vrrp[3954]: VRRP_Instance(VI_1)Entering MASTER STATE
Sep 25 08:15:22 localhost Keepalived_vrrp[3954]: VRRP_Instance(VI_1)setting protocol VIPs.
由於沒人和主機2搶，所以主機2就升級爲master
# ip addrshow
(1): lo:<LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd00:00:00:00:00:00
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 qlen1000
link/ether 00:0c:29:c8:8f:e1 brdff:ff:ff:ff:ff:ff
inet 172.16.31.30/16 brd 172.16.255.255scope global eth0
inet172.16.31.22/32 scope global eth0 此刻ip在主機2上
inet6 fe80::20c:29ff:fec8:8fe1/64 scopelink
valid_lftforever preferred_lft forever
7， 在2節點上停止服務，在開始一下服務
Oct 10 17:40:59 localhost Keepalived_vrrp[2941]: VRRP_Instance(VI_1)Received lower prio advert, forcing new election 1節點接收到一個比自己優先級低的，然後重新選舉，自己變爲主
Oct 10 17:41:00 localhost Keepalived_vrrp[2941]: VRRP_Instance(VI_1)Entering MASTER STATE
Oct 10 17:41:00 localhost Keepalived_vrrp[2941]: VRRP_Instance(VI_1)setting protocol VIPs.
Oct 10 17:41:00 localhost Keepalived_healthcheckers[2940]: Netlinkreflector reports IP 172.16.31.22 added
8， 然後開始配置雙主模型，在1,2節點上配置，各增加一個實例
vrrp_instanceVI_2 {
state BACKUP
interface eth0
virtual_router_id 200
priority 99
advert_int 1
authentication {
auth_type PASS
auth_pass 123
}
virtual_ipaddress {
172.16.31.222
}
}
節點2上
vrrp_instanceVI_2 {
state MASTER
interface eth0
virtual_router_id 200
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass 123
}
virtual_ipaddress {
172.16.31.222
}
}然後節點1上# ip addr show
（1）: lo: <LOOPBACK,UP,LOWER_UP> mtu16436 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 qlen1000
link/ether 00:0c:29:a7:90:c5 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.10/16 brd 172.16.255.255 scope global eth0
inet 172.16.31.22/32 scope global eth0
inet6 fe80::20c:29ff:fea7:90c5/64 scope link
valid_lft forever preferred_lft forever
節點2上
（2）: eth0:<BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen1000
link/ether 00:0c:29:c8:8f:e1 brd ff:ff:ff:ff:ff:ff
inet 172.16.31.30/16 brd 172.16.255.255 scope global eth0
inet 172.16.31.222/32 scope global eth0
inet6 fe80::20c:29ff:fec8:8fe1/64 scope link
valid_lft forever preferred_lft forever
9， 簡單的利用keepalived配置通知nginx
首先在1,2節點上安裝nginx，然後再配置文件keepalived.conf配置
vrrp_instanceVI_1 {
state MASTER
interface eth0
virtual_router_id 101
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass 123456
}
virtual_ipaddress {
172.16.31.22
}
notify_master "/etc/rc.d/init.d/nginxstart" 表示當切換master時，要執行的腳本
notify_backup "/etc/rc.d/init.d/nginxstop"
notify_fault "/etc/rc.d/init.d/nginxstop"
}2節點上同樣的配置，然後1,2節點
#servicekeepalived restart
LISTEN 0 128 *:80 *:*
然後再此基礎上擴展，shell腳本來檢測
vrrp-script chk_nx{
script "killall -0 nginx"
interval 1 腳本執行間隔
weight -5 腳本結果導致優先級變更，-5優先級-5，+10優先級+10
fall 2
rise 1
}
然後再配置文件最後再加上
track_script{
chk_ nx 這個是調用上面那個腳本
}2節點上也是這樣，然後2節點開啓服務
然後開啓1節點，這時候由於1節點比2節點的優先級高，就會變爲主的
