一、集群高可用概述
單純的lvs/nginx反向代理模型做負載集群應用時,DR(director)存在單點故障隱患,故需要有機制來保證DR的高可用性。常用的高可用性方案有Keepalived、corosync,Keepalived主要是由VRRP協議實現了VIPfloating,比較適用于前端DR的高可用性,Corosync一般用于更專業的集群模型實現Service的高可用。Keepalived起初就是為了實現LVS集群director高可用而開發的,本處僅做Keepalived+LVS-DR模型實驗。
Keepalived中優先級高的節點為MASTER。MASTER其中一個職責就是響應VIP的arp包,將VIP和mac地址映射關系告訴局域網內其 他主機,同時,它還會以多播的形式向局域網中發送VRRP通告,告知BACKUP組自己的優先級。網絡中的所有BACKUP節點只負責 處理MASTER發出的多播包,當發現MASTER的優先級沒自己高(腳本檢測故障觸發自我降級),或者沒收到MASTER的VRRP通告(網絡故障/MASTER宕機)時,BACKUP將自己切換到MASTER狀 態,然后做MASTER該做的事:1.響應arp包,2.發送VRRP通告。
三、實驗環境
1.網絡拓補圖
?2.軟件環境
-
-
keepalived.x86_64???????????????????????? 1.3.5-6.el7
global_defs { ? ? ? ? ? ? ? ? ? notification_email { ? ? ? ? ? ? ? ? ? ? ? root@localhost #此處僅發給本機,更定制化的郵件通知功能一般由zabbix來做。 ? ? ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? ? notification_email_from keepalived@localhost ? ? ? ? ? ? ? ? ? smtp_server 127.0.0.1 ? ? ? ? ? ? ? ? ? smtp_connect_timeout 30 ? ? ? ? ? ? ? ? ? router_id keepalivedR1 ?#定義路由器標識,每臺服務器局域網內唯一就行。 ? ? ? ? ? ? ? ? ? vrrp_mcast_group4 224.0.0.33 #定義master向backup組播vrrp報文的地址。 ? ? ? ? ? ? ? } ? ? vrrp_script chk_down { ? ? ? ? ? ? ? script "/etc/keepalived/chk_down.sh" ? ? ? ? ? ? ? interval 1 ? #腳本檢測間隔 ? ? ? ? ? ? ? weight -15 ? #即原有優先級+weight,負值即降低。 # 注意當weight=0時, 用于變更vrrp_instance的狀態例如腳本檢測失敗, 則vrrp的狀態直接變為FAULT. (不管有沒有其他節點存在來接管MASTER) ? ? ? ? ? ? ? fall 2 ? ? ? #執行腳本兩次exit非0則降低優先級 ? ? ? ? ? ? ? rise 1 #執行腳本一次exit 0 則還原優先級 ? ? ? ? ? ? ? user keepalived_script #默認用戶yum安裝keepalived需自建,不存在則調用root(不推薦) ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? vrrp_instance VI_1 { ? ? ? ? ? ? ? ? ? state MASTER ? ?#自定義的state,但如果你的優先級小于backup也不會成為master。 ? ? ? ? ? ? ? ? ? interface ens39 ?#綁定為當前虛擬路由器使用的物理接口。 ? ? ? ? ? ? ? ? ? virtual_router_id 3 ?#當前虛擬路由器的惟一標識,范圍是0-255。同實例一致! ? ? ? ? ? ? ? ? ? priority 99 #初始優先級,范圍1-254。 ? ? ? ? ? ? ? ? ? advert_int 1 #vrrp通告的時間間隔。 ? ? ? ? ? ? ? ? ? authentication { ? ? ? ? ? ? ? ? ? ? ? auth_type PASS ? #簡單密碼驗證,不超過8位。 ? ? ? ? ? ? ? ? ? ? ? auth_pass 736w4ib2 #最好使用隨機字符串,同vip實例保持一致! ? ? ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? ? virtual_ipaddress { ? ? ? ? ? ? ? ? ? ? ? ?192.168.7.120/24 dev ens39 ? ? ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? ? notify_master "/etc/keepalived/notify.sh master" #調用通知腳本 ? ? ? ? ? ? ? ? ? notify_backup "/etc/keepalived/notify.sh backup" ? ? ? ? ? ? ? ? ? notify_fault "/etc/keepalived/notify.sh fault" ? ? ? ? ? ? ? } ? virtual_server 192.168.7.120 80 { ?#此處可用IP port/fwmark id 標識VIP對應服務 ? ? ? ? ? ? ? ? ? ? delay_loop 2 ? #服務輪詢的時間間隔2s ? ? ? ? ? lb_algo rr ? ? ? ? ? lb_kind DR ? ? ? ? ? protocol TCP ? ? ? ? ? ? ? ? ? real_server 192.168.7.125 80 { ? ? ? ? ? ? ? ? ? ? ? weight 1 ? ? ? ? ? ? ? ? ? ? ? HTTP_GET { ? ? ? ? ? ? ? ? ? ? ? url { ? ? ? ? ? ? ? ? ? ? ? ? ? path / ? ? ? ? ? ? ? ? ? ? ? ? ? status_code 200 ? ? ? ? ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? ? ? ? connect_timeout 2 ? ? ? ? ? ? ? ? ? ? ? nb_get_retry 3 ? ? ? ? ? ? ? ? ? ? ? delay_before_retry 3 ? ? ? ? ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? ? real_server 192.168.7.126 80 { ? ? ? ? ? ? ? ? ? ? ? weight 1 ? ? ? ? ? ? ? ? ? ? ? HTTP_GET { ? ? ? ? ? ? ? ? ? ? ? ? ? url { ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? path / ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? status_code 200 ? ? ? ? ? ? ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? ? ? ? ? ? connect_timeout 2 ? ? ? ? ? ? ? ? ? ? ? ? ? nb_get_retry 3 ? ? ? ? ? ? ? ? ? ? ? ? ? delay_before_retry 3 ? ? ? ? ? ? ? ? ? ? ? ? ? } ? ? ? ? ? ? ? ? }
將上述配置錄入/etc/keepalived/keepalived.conf中,過程如下:
[root@DR1 ~]# cd /etc/keepalived/ [root@DR1 keepalived]# cp keepalived.conf{,.bak} #備份下配置文件 [root@DR1 keepalived]# ls keepalived.conf keepalived.conf.bak [root@DR1 keepalived]# vim keepalived.conf ?
-
配置中調用的notify腳本內容如下:
? ?#!/bin/bash ? ?# ? ?contact='root@localhost' ? ? notify() { ? ? ? local mailsubject="$(hostname) to be $1, vip floating" ? ? ? local mailbody="$(date +'%F %T'): vrrp transition, $(hostname) changed to be $1" ? ? ? ?echo "$mailbody" | mail -s "$mailsubject" $contact ? } ? ? case $1 in ? master) ? ? ? notify master ? ? ? ;; ? backup) ? ? ? notify backup ? ? ? ;; ? fault) ? ? ? notify fault ? ? ? ;; ? *) ? ? ? ?echo "Usage: $(basename $0) {master|backup|fault}" ? ? ? ?exit 1 ? ? ? ;; ? ?esac ? ? ? ? ? ?
-
chk_down.sh只有條件判斷:
[[ -f /etc/keepalived/down ]] && exit 1 || exit 0
-
配置DR2
DR2,僅需修改如下配置:
-
router_id keepalivedR2
-
state BACKUP
-
priority 90
(2)兩臺RS配置
兩臺RS均yum安裝nginx,啟動服務監聽80端口即可。
需要注意的有如下幾點:
-
檢查系統是否自帶httpd服務并停止,否則會與nginx發生沖突。
-
添加vip到網卡輔助接口,并修改arp參數抑制apr報文響應,此處可用腳本如下:
vip=192.168.7.120 /usr/sbin/ip addr add $vip/32 dev lo label lo:0 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 ?
注意:因客戶端在同網段,添加完RS抑制arp響應后,需要將客戶端的arp緩存清空防止干擾效果。
-
修改默認主頁,主要是做標記區分RS1/RS2??刹捎萌缦路椒ǎ?/span>
編輯默認index.html 在vim末行模式輸入:%s/nginx/server1/g回車即可
五、測試實驗效果
-
測試VIP漂移
-
先在DR2初始BACKUP開啟keepalived觀察是否有狀態變更:
[root@DR2 ~]# systemctl start keepalived.service [root@DR2 ~]# systemctl status keepalived.service ?.keepalived.service - LVS and VRRP High Availability Monitor ? Loaded: loaded (/usr/lib/systemd/system/keepalived.service; disabled; vendor preset: disabled) ? Active: active (running) since Sun 2018-07-15 13:28:10 CST; 19s ago Process: 2620 ExecStart=/usr/sbin/keepalived $KEEPALIVED_OPTIONS (code=exited, status=0/SUCCESS) Main PID: 2621 (keepalived) ? CGroup: /system.slice/keepalived.service ? ? ? ? ? ?..2621 /usr/sbin/keepalived -D ? ? ? ? ? ?..2622 /usr/sbin/keepalived -D ? ? ? ? ? ?..2623 /usr/sbin/keepalived -D ? Jul 15 13:28:14 DR2 Keepalived_vrrp[2623]: Sending gratuitous ARP on ens39 for 192.168.7.120 Jul 15 13:28:14 DR2 Keepalived_vrrp[2623]: Sending gratuitous ARP on ens39 for 192.168.7.120 Jul 15 13:28:14 DR2 Keepalived_vrrp[2623]: Sending gratuitous ARP on ens39 for 192.168.7.120 Jul 15 13:28:14 DR2 Keepalived_vrrp[2623]: Opening script file /etc/keepalived/notify.sh Jul 15 13:28:19 DR2 Keepalived_vrrp[2623]: Sending gratuitous ARP on ens39 for 192.168.7.120 Jul 15 13:28:19 DR2 Keepalived_vrrp[2623]: VRRP_Instance(VI_1) Sending/queueing gratuitous ARPs on ens39 for 192.168.7.120 [root@DR2 ~]# mail Heirloom Mail version 12.5 7/5/10. Type ? for help. "/var/spool/mail/root": 6 messages 2 new 2 unread >N ?5 root ? ? ? ? ? ? ? ? Sun Jul 15 13:28 ?18/667 ? "DR2 to be backup, vip floating" N ?6 root ? ? ? ? ? ? ? ? Sun Jul 15 13:28 ?18/667 ? "DR2 to be master, vip floating" & 6 Message ?6: From root@DR2.localdomain Sun Jul 15 13:28:15 2018 Return-Path: <root@DR2.localdomain> X-Original-To: root@localhost Delivered-To: root@localhost.localdomain Date: Sun, 15 Jul 2018 13:28:14 +0800 To: root@localhost.localdomain Subject: DR2 to be master, vip floating User-Agent: Heirloom mailx 12.5 7/5/10 Content-Type: text/plain; charset=us-ascii From: root@DR2.localdomain (root) Status: R ? 2018-07-15 13:28:14: vrrp transition, DR2 changed to be master
可以看到,因為master沒上線,DR2由backup變為master。我們的notify.sh腳本成功發送狀態變更通知到本地郵箱。
-
測試DR1上線效果:
[root@DR2 ~]# systemctl status keepalived.service Jul 15 14:52:48 DR2 Keepalived_vrrp[1812]: VRRP_Instance(VI_1) Received advert with higher priority 99, ours 90 Jul 15 14:52:48 DR2 Keepalived_vrrp[1812]: VRRP_Instance(VI_1) Entering BACKUP STATE Jul 15 14:52:48 DR2 Keepalived_vrrp[1812]: VRRP_Instance(VI_1) removing protocol VIPs. Jul 15 14:52:48 DR2 Keepalived_vrrp[1812]: Opening script file /etc/keepalived/notify.sh You have new mail in /var/spool/mail/root [root@DR2 ~]# ip addr show 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1 ? link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 ? inet 127.0.0.1/8 scope host lo ? ? ? valid_lft forever preferred_lft forever ? inet6 ::1/128 scope host ? ? ? valid_lft forever preferred_lft forever 2: ens39: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000 ? link/ether 00:50:56:3b:a3:7c brd ff:ff:ff:ff:ff:ff ? inet 192.168.7.122/24 brd 192.168.7.255 scope global ens39 ? ? ? valid_lft forever preferred_lft forever ? inet6 fe80::250:56ff:fe3b:a37c/64 scope link ? ? ? valid_lft forever preferred_lft forever [root@DR1 ~]# systemctl status keepalived.service Jul 15 14:52:49 DR1 Keepalived_vrrp[3471]: Sending gratuitous ARP on ens39 for 192.168.7.120 Jul 15 14:52:49 DR1 Keepalived_vrrp[3471]: Opening script file /etc/keepalived/notify.sh Jul 15 14:52:54 DR1 Keepalived_vrrp[3471]: Sending gratuitous ARP on ens39 for 192.168.7.120 Jul 15 14:52:54 DR1 Keepalived_vrrp[3471]: VRRP_Instance(VI_1) Sending/queueing gratuitous ARPs on ens39 for 192.168.7.120
DR2因為收到了更高優先級的VRRP組播報文而卸載VIP,DR1獲取到VIP。
-
手工touch生成/etc/keepalived/down文件模擬單機故障也會發生上述VIP漂移現象。
Jul 15 15:21:51 DR1 Keepalived_vrrp[3907]: WARNING - default user 'keepalived_script' for script execution does not exist - please create. # 此處踩坑了,提示conf文件中配置的腳本需要keepalived_script用戶執行 # 還有個巨坑是參考資料中直接在keepalived.conf中定義script "[[ -f /etc/keepalived/down ]] && # exit 1 || exit 0" 但是我多次測試不成功,將條件判斷放入腳本,conf文件僅引用路徑才成功...... Jul 15 15:54:24 DR1 Keepalived_vrrp[4438]: /etc/keepalived/chk_down.sh exited with status 1 Jul 15 15:54:24 DR1 Keepalived_vrrp[4438]: VRRP_Script(chk_down) failed Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: VRRP_Instance(VI_1) Changing effective priority from 99 to 84 Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: /etc/keepalived/chk_down.sh exited with status 1 Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: VRRP_Instance(VI_1) Received advert with higher priority 90, ours 84 Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: VRRP_Instance(VI_1) Entering BACKUP STATE Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: VRRP_Instance(VI_1) removing protocol VIPs. Jul 15 15:54:25 DR1 Keepalived_vrrp[4438]: Opening script file /etc/keepalived/notify.sh
-
-
測試RS故障檢測
-
先測試RS是否正常被輪詢
for i in {1,5};do curl http://192.168.7.120 ;done
-
手動關停一臺rs的nginx
[root@RS2 ~]# systemctl stop nginx [root@DR1 ~]# journalctl -xe Jul 15 16:59:04 DR1 Keepalived_healthcheckers[4437]: Error connecting server [192.168.7.126]:80. [root@DR1 ~]# systemctl status keepalived Jul 15 16:59:05 DR1 Keepalived_healthcheckers[4437]: Error connecting server [192.168.7.126]:80. Jul 15 16:59:06 DR1 Keepalived_healthcheckers[4437]: Error connecting server [192.168.7.126]:80. Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: Error connecting server [192.168.7.126]:80. Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: Check on service [192.168.7.126]:80 failed after 3 retry. Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: Removing service [192.168.7.126]:80 from VS [192.168.7.120]:80 Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: Remote SMTP server [127.0.0.1]:25 connected. Jul 15 16:59:07 DR1 Keepalived_healthcheckers[4437]: SMTP alert successfully sent. [root@DR1 ~]# mail N 16 keepalived@localhost Sun Jul 15 16:59 ?17/646 ? "[keepalivedR1] Realserver [192.168.7.126]:80 - DOWN"
-
重新啟動nginx
Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: HTTP status code success to [192.168.7.126]:80 url(1). Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: Remote Web server [192.168.7.126]:80 succeed on service. Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: Adding service [192.168.7.126]:80 to VS [192.168.7.120]:80 Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: Remote SMTP server [127.0.0.1]:25 connected. Jul 15 17:02:55 DR1 Keepalived_healthcheckers[4437]: SMTP alert successfully sent. You have new mail in /var/spool/mail/root
-
-
本文來自投稿,不代表Linux運維部落立場,如若轉載,請注明出處:http://www.www58058.com/103020