基于corosync+pacemaker实现nfs+nginx部署

dfsfd

pcs相关配置：（因为在7版本，所以pcs支持比较好，crmsh比较复杂）

环境主机-centos7：node1：172.25.0.29 node2：172.25.0.30

配置集群的前提：

1、时间同步

2、主机名互相访问

3、是否使用仲裁设备。

生命周期管理工具主要包括以下：

Pcs:agent(pcsd) ：应用于corosync+pacemaker

Crash:pssh ： 应用于ansible相关的服务

一、安装corosync+pacemaker和crm管理包

1、先配置相关主机和相关时间同步服务器：

node1：

1 2 3 4 5 6 7

[iyunv@node1 ~]# cat /etc/hosts 127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4 ::1         localhost localhost.localdomain localhost6 localhost6.localdomain6 172.25.0.29 node1 172.25.0.30 node2 [iyunv@node1 ~]# crontab -e */5 * * * * ntpdate cn.pool.ntp.org   ###添加任务

node2：

1 2 3 4 5 6 7

[iyunv@node2 ~]# cat /etc/hosts 127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4 ::1         localhost localhost.localdomain localhost6 localhost6.localdomain6 172.25.0.29 node1 172.25.0.30 node2 [iyunv@node1 ~]# crontab -e */5 * * * * ntpdate cn.pool.ntp.org   ###添加任务

在node1和node2上可以看到已经添加时间任务：

1 2 3 4

[iyunv@node1 ~]# crontab -l */5 * * * * ntpdate cn.pool.ntp.org [iyunv@node2 ~]# crontab -l */5 * * * * ntpdate cn.pool.ntp.org

添加node1和node2的信任关系

1 2 3 4 5 6 7

[iyunv@node1 ~]# ssh-keygen [iyunv@node1 ~]# ssh-copy-id node2 The authenticity of host 'node2 (172.25.0.30)' can't be established. ECDSA key fingerprint is ae:88:02:59:f9:7f:e9:4f:48:8d:78:d2:6f:c7:7a:f1. Are you sure you want to continue connecting (yes/no)? yes /usr/bin/ssh-copy-id: INFO: attempting to log in with the new key(s), to filter out any that are already installed /usr/bin/ssh-copy-id: WARNING: All keys were skipped because they already exist on the remote system.

我这里已经添加了,才会出现警告

2、在node1和node2个结点上执行：

1 2	[iyunv@node1 corosync]# yum install -y pacemaker pcs psmisc policycoreutils-python [iyunv@node2 corosync]# yum install -y pacemaker pcs psmisc policycoreutils-python

3、node1和node2上启动pcs并且让开机启动：

1 2 3 4

[iyunv@node1 corosync]# systemctl start pcsd.service [iyunv@node1 corosync]# systemctl enable pcsd [iyunv@node2 corosync]# systemctl start pcsd.service [iyunv@node2 corosync]# systemctl enable pcsd

4、在两台主机上修改用户hacluster的密码:

1 2	[iyunv@node1 corosync]# echo 123456 | passwd --stdin hacluster [iyunv@node2 corosync]# echo 123456 | passwd --stdin hacluster

下面的可以一台主机同步配置了

node1上：

5、注册pcs集群主机（默认注册使用用户名hacluster，和密码）：

1 2 3

[iyunv@node1 corosync]# pcs cluster auth node1 node2     ##设置注册那个集群节点 node2: Already authorized node1: Already authorized

6、在集群上注册两台集群：

1	[iyunv@node1 corosync]# pcs cluster setup --name mycluster node1 node2 --force   ##设置集群

7、接下来就在某个节点上已经生成来corosync配置文件：

1 2 3

[iyunv@node1 ~]# cd /etc/corosync/  ##进入corosync目录 [iyunv@node1 corosync]# ls corosync.conf  corosync.conf.example  corosync.conf.example.udpu  corosync.xml.example  uidgid.d

#我们看到生成来corosync.conf配置文件：

8、启动集群：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

[iyunv@node2 corosync]# pcs cluster start --all node1: Starting Cluster... node2: Starting Cluster... ##相当于启动pacemaker和corosync: [iyunv@node1 corosync]#  ps -ef | grep corosync root      19586      1  0 18:05 ?        00:00:40 corosync root      29230  21295  0 19:13 pts/1    00:00:00 grep --color=auto corosync [iyunv@node1 corosync]# ps -ef | grep pacemaker root       1843      1  0 11:21 ?        00:00:04 /usr/libexec/pacemaker/lrmd haclust+   1845      1  0 11:21 ?        00:00:03 /usr/libexec/pacemaker/pengine root      19593      1  0 18:05 ?        00:00:01 /usr/sbin/pacemakerd -f haclust+  19594  19593  0 18:05 ?        00:00:01 /usr/libexec/pacemaker/cib root      19595  19593  0 18:05 ?        00:00:00 /usr/libexec/pacemaker/stonithd haclust+  19596  19593  0 18:05 ?        00:00:00 /usr/libexec/pacemaker/attrd haclust+  19597  19593  0 18:05 ?        00:00:01 /usr/libexec/pacemaker/crmd root      29288  21295  0 19:14 pts/1    00:00:00 grep --color=auto pacemaker ###可以看到corosync和pacemaker已经起来了

9、查看集群的状态

1 2 3 4 5 6 7 8 9 10 11 12 13

[iyunv@node1 corosync]# corosync-cfgtool -s Printing ring status. Local node ID 1 RING ID 0 id= 172.25.0.29 status= ring 0 active with no faults [iyunv@node1 corosync]# ssh node2 corosync-cfgtool -s Printing ring status. Local node ID 2 RING ID 0 id= 172.25.0.30 status= ring 0 active with no faults ###可以发现node1和node2的集群都已经起来。

10、到这里我们先查看集群是否有错：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

[iyunv@node1 corosync]# crm_verify -L -V    error: unpack_resources:     Resource start-up disabled since no STONITH resources have been defined    error: unpack_resources:     Either configure some or disable STONITH with the stonith-enabled option    error: unpack_resources:     NOTE: Clusters with shared data need STONITH to ensure data integrity Errors found during check: config not valid ##发现有错，要我们关掉 stonith-enabled，避免下一步出错我们先关掉这个 [iyunv@node1 corosync]# pcs property set stonith-enabled=false [iyunv@node1 corosync]# crm_verify -L -V [iyunv@node1 corosync]# pcs property list Cluster Properties: cluster-infrastructure: corosync cluster-name: mycluster dc-version: 1.1.16-12.el7_4.2-94ff4df have-watchdog: false stonith-enabled: false

11、现在我们可以下载安装crmsh来操作(从github来下载，然后解压直接安装)：

https://codeload.github.com/ClusterLabs/crmsh/tar.gz/2.3.2

node1上：

1 2 3 4 5 6 7 8

[iyunv@node1 ~]# cd /usr/local/src/ [iyunv@node1 src]# ls crmsh-2.3.2.tar       [iyunv@node1 src]#tar xvf crmsh-2.3.2.tar [iyunv@node1 src]# ls crmsh-2.3.2.tar crmsh-2.3.2 [iyunv@node1 src]# cd crmsh-2.3.2 [iyunv@node1 crmsh-2.3.2]# python setup.py install  ##编译安装

node2上：跟node1同样的操作

二、源代码安装nginx和安装nfs

###在node1和node2安装nginx，下面是node1的操作：

1、安装nginx软件依赖包：

1 2	yum -y groupinstall "Development Tools" "Server Platform Deveopment" yum -y install openssl-devel pcre-devel

2、在所有的主机上面都操作，下载nginx包

1	[iyunv@node1 src]# yum install wget –y               ##安装wget工具

3、下载nginx包：

1	[iyunv@node1 src]# wget http://nginx.org/download/nginx-1.12.0.tar.gz

4、添加nginx运行的用户：

1	[iyunv@node1 sbin]# useradd nginx

5解压nginx包，并且安装：

1 2	[iyunv@node1 src]# tar zxvf nginx-1.12.0.tar.gz [iyunv@node1 src]# cd nginx-1.12.0/

6、安装nginx包：

1 2 3 4

[iyunv@node1 nginx-1.12.0]# ./configure --prefix=/usr/local/nginx --user=nginx --group=nginx --with-http_ssl_module --with-http_flv_module --with-http_stub_status_module --with-http_gzip_static_module  --with-pcre ###编译安装 [iyunv@node1 nginx-1.12.0]# make && make install node1、node2装完后测试nginx

6、测试nginx：

node1上：

1 2 3

[iyunv@node1 nginx]# cd /usr/local/nginx/ [iyunv@node1 nginx]# echo node1 >　html/index.html [iyunv@node1 nginx]#/usr/local/nginx/sbin/nginx

node2上：

1 2 3

[iyunv@node2 nginx]# cd /usr/local/nginx/ [iyunv@node2 nginx]# echo node2 >　html/index.html [iyunv@node2 nginx]#/usr/local/nginx/sbin/nginx

访问web服务：

1 2 3 4

[iyunv@node1 nginx]#curl 172.25.0.29 node1 [iyunv@node1 nginx]#curl 172.25.0.29 node2

node1、node2可以正常访问

把nginx关闭，因为等会利用corosync和pacemaker自动管理nginx

建个nginx启动脚本，等下启动nginx需要，在node1和node2上都要新建

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

[iyunv@node1 ~]# cat /etc/systemd/system/nginx.service [Unit] Description=nginx After=network.target     [Service] Type=forking ExecStart=/usr/local/nginx/sbin/nginx ExecReload=/usr/local/nginx/sbin/nginx -s reload ExecStop=/usr/local/nginx/sbin/nginx -s quit PrivateTmp=true     [Install] WantedBy=multi-user.target ##node2上也同样的操作

需要给脚本执行权限

1 2 3 4

[iyunv@node1 ~]# chmod a+x /etc/systemd/system/nginx.service [iyunv@node2 ~]# chmod a+x /etc/systemd/system/nginx.service [iyunv@node1 ~]# systemctl enable nginx [iyunv@node1 ~]# systemctl enable nginx   ##在systemd资源代理下，要有enable 才能被crm识别，所以要把nginx enable掉

nfs搭建：

nfs的作用我们都明确，所以我们只需在一台上安装就好，我这里在node1安装

1 2 3 4 5 6 7 8 9

[iyunv@node1 ~]#yum install -y rpc-bind nfs-utils [iyunv@node1 ~]# mkdir /www   ###新建www的目录，等会用于共享。 [iyunv@node1 ~]# cat /etc/exports /www  *(rw,async,no_root_squash) [iyunv@node1 ~]#systemctl restart nfs     ###重启nfs [iyunv@node1 ~]# showmount -e 172.25.0.29 Export list for 172.25.0.29: /www *                     ##可以发现www这个目录已经共享了 [iyunv@node1 ~]# echo node  >　 /www/index.html    ###给共享目录添加index.html,用于虚拟ip的访问

三、高可用实现nfs+nginx

1、资源嗲里的使用方法：

在node1上配置：

1 2 3 4 5 6 7 8 9

[iyunv@node1 ~]# crm ra crm(live)ra# info systemd:nginx systemd unit file for nginx (systemd:nginx) Cluster Controlled nginx Operations' defaults (advisory minimum):     start         timeout=100     stop          timeout=100     status        timeout=100     monitor       timeout=100 interval=60

2、进入配置模式configure下：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

crm(live)ra# cd crm(live)#cd configure crm(live)configure# primitive webip ocf:heartbeat:IPaddr params ip=172.25.0.100  ###添加虚拟ip ##配置好之后用show查看 crm(live)configure# show node 1: node1 node 2: node2 primitive webip IPaddr \         params ip=172.25.0.100 property cib-bootstrap-options: \         have-watchdog=false \         dc-version=1.1.16-12.el7_4.2-94ff4df \         cluster-infrastructure=corosync \         cluster-name=mycluster \         stonith-enabled=false crm(live)configure# verify      #检查脚本是否有错 crm(live)configure# commit      ##提交、保存 crm(live)configure# cd

3、定义web服务资源：

进入配置模式configure：

1 2 3 4 5 6 7 8

crm(live)configure# primitive webserver systemd:nginx      ##添加nginx服务 crm(live)configure# verify WARNING: webserver: default timeout 20s for start is smaller than the advised 100 WARNING: webserver: default timeout 20s for stop is smaller than the advised 100 ### 小于时间间隔会有警告，可以不用理会。 crm(live)configure# commit   WARNING: webserver: default timeout 20s for start is smaller than the advised 100 WARNING: webserver: default timeout 20s for stop is smaller than the advised 100

##提交有个警告不用管：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

crm(live)configure# show node 1: node1 \         attributes standby=off node 2: node2 primitive vip IPaddr \         params ip=172.25.0 primitive web systemd:nginx \         op monitor interval=30s timeout=100s \         op start timeout=100s interval=0 \         op stop timeout=100s interval=0 property cib-bootstrap-options: \         have-watchdog=false \         dc-version=1.1.16-12.el7_4.4-94ff4df \         cluster-infrastructure=corosync \         cluster-name=mycluster \         stonith-enabled=false

##我们检测下已经有两个资源了：

1 2 3 4 5 6 7 8 9 10 11 12

crm(live)configure# cd crm(live)# status Stack: corosync Current DC: node1 (version 1.1.16-12.el7_4.2-94ff4df) - partition with quorum Last updated: Sat Oct 14 21:20:59 2017 Last change: Sat Oct 14 21:17:43 2017 by root via cibadmin on node1 2 nodes configured 2 resources configured Online: [ node1 node2 ] Full list of resources: webip  (ocf::heartbeat:IPaddr):        Started node2 webserver      (systemd:nginx):        Started node1

##我们也发现默认资源也是均衡了，但是我们发现不均衡了分配了资源，但是我们需要定义是一个组的，所以把两个资源加一组(为了实现高可用)

把两个添加到同个组里面：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

crm(live)# configure crm(live)configure# group webservice webip webserver   ##添加 webservice webip在同个组里面 crm(live)configure# verify crm(live)configure# commit crm(live)configure# cd .. crm(live)# status Stack: corosync Current DC: node1 (version 1.1.16-12.el7_4.2-94ff4df) - partition with quorum Last updated: Sat Oct 14 21:24:17 2017 Last change: Sat Oct 14 21:24:12 2017 by root via cibadmin on node1 2 nodes configured 2 resources configured Online: [ node1 node2 ] Full list of resources: Resource Group: webservice      webip      (ocf::heartbeat:IPaddr):        Started node1      webserver  (systemd:httpd):        Started node1                 ##可以发现 webservice webip在同个组里面了

4、定义nfs资源：

查看文件系统类型

1 2 3 4 5 6 7

crm(live)ra# info ocf:heartbeat:Filesystem device* (string): block device     The name of block device for the filesystem, or -U, -L options for mount, or NFS mount specification. directory* (string): mount point     The mount point for the filesystem. fstype* (string): filesystem type     The type of filesystem to be mounted.

###有三个必填项目

##开始配置

1	crm(live)configure# primitive webstore ocf:heartbeat:Filesystem params device="172.25.0.29:/www" directory="/usr/local/nginx/html" fstype="nfs" op start timeout=60s op stop timeout=60s op monitor interval=20s timeout=40s      ###定义/www 挂载到/usr/local/nginx/html下

5、定义排列约束：

1 2 3 4 5

crm(live)configure# colocation webserver_with_webstore_and_webip inf: webserver ( webip webstore) crm(live)configure# verify WARNING: webserver_with_webstore_and_webip: resource webserver is grouped, constraints should apply to the group WARNING: webserver_with_webstore_and_webip: resource webip is grouped, constraints should apply to the group crm(live)configure# commit

##查看状态：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

crm(live)configure# show node 1: node1 \ attributes standby=off node 2: node2 primitive webip IPaddr \ params ip=172.25.0.100 primitive webserver systemd:nginx \ op monitor interval=30s timeout=100s \ op start timeout=60s interval=0 \ op stop timeout=60s interval=0 primitive webstore Filesystem \ params device="172.25.0.29:/www" directory="/usr/local/nginx/html" fstype=nfs \ op start timeout=60s interval=0 \ op stop timeout=60s interval=0 \ op monitor interval=20s timeout=40s group webservice webip webserver colocation webserver_with_webstore_and_webip inf: webserver ( webip webstore ) property cib-bootstrap-options: \ have-watchdog=false \ dc-version=1.1.16-12.el7_4.4-94ff4df \ cluster-infrastructure=corosync \ cluster-name=mycluster \ stonith-enabled=false \

6、定义执行顺序：

1 2 3 4

crm(live)configure# order webstore_after_webip Mandatory: webip webstore crm(live)configure# verify crm(live)configure# order webserver_after_webstore Mandatory: webstore webserver crm(live)configure#

###查看一下状态

1 2 3 4 5 6 7 8 9 10 11 12 13 14

crm(live)# status Stack: corosync Current DC: node1 (version 1.1.16-12.el7_4.4-94ff4df) - partition with quorum Last updated: Wed Oct 25 20:46:41 2017 Last change: Wed Oct 25 16:56:52 2017 by root via cibadmin on node1 2 nodes configured 3 resources configured Online: [ node1 node2 ] Full list of resources: Resource Group: webservice      webip(ocf::heartbeat:IPaddr):Started node1      webserver(systemd:nginx):Started node1 webstore(ocf::heartbeat:Filesystem):Started node1 ##可以看到我们的顺序是webip webserver webstore

7、测试

1 2 3 4 5 6 7 8 9 10 11 12 13 14

[iyunv@node1 ~]# ip addr 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: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000     link/ether 00:0c:29:49:e9:da brd ff:ff:ff:ff:ff:ff     inet 172.25.0.29/24 brd 172.25.0.255 scope global ens33        valid_lft forever preferred_lft forever     inet 172.25.0.100/24 brd 172.25.0.255 scope global secondary ens33        valid_lft forever preferred_lft forever     inet6 fe80::20c:29ff:fe49:e9da/64 scope link

可以看到vip已经起来了

接下来访问web服务：

1 2	[iyunv@node1 ~]# curl 172.25.0.100 node

可以发现访问的是/www/index里的内用

我们把node1的pacemaker和corosync停掉

1 2	[iyunv@node1 ~]# systemctl stop pacemaker    ##先关pacemaker先 [iyunv@node1 ~]# systemctl stop corosync

在node2上可以看到node2已经接管了

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

[iyunv@node2 crmsh-2.3.2]# crm crm(live)# status Stack: corosync Current DC: node2 (version 1.1.16-12.el7_4.4-94ff4df) - partition with quorum Last updated: Wed Oct 25 20:54:33 2017 Last change: Wed Oct 25 16:56:52 2017 by root via cibadmin on node1 2 nodes configured 3 resources configured Online: [ node2 ] OFFLINE: [ node1 ] Full list of resources: Resource Group: webservice      webip(ocf::heartbeat:IPaddr):Started node2      webserver(systemd:nginx):Started node2 webstore(ocf::heartbeat:Filesystem):Started node2 crm(live)#exit [iyunv@node2 crmsh-2.3.2]# ip addr 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: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000     link/ether 00:0c:29:64:00:b1 brd ff:ff:ff:ff:ff:ff     inet 172.25.0.30/24 brd 172.25.0.255 scope global ens33        valid_lft forever preferred_lft forever     inet 172.25.0.100/24 brd 172.25.0.255 scope global secondary ens33        valid_lft forever preferred_lft forever     inet6 fe80::20c:29ff:fe64:b1/64 scope link

##vip已经转移到node2上

1 2 3 4 5 6 7 8 9 10

[iyunv@node2 crmsh-2.3.2]# df -h Filesystem           Size  Used Avail Use% Mounted on /dev/mapper/cl-root   18G  2.5G   16G  14% / devtmpfs             226M     0  226M   0% /dev tmpfs                237M   86M  151M  37% /dev/shm tmpfs                237M  8.6M  228M   4% /run tmpfs                237M     0  237M   0% /sys/fs/cgroup /dev/sda1           1014M  197M  818M  20% /boot tmpfs                 48M     0   48M   0% /run/user/0 172.25.0.29:/www      18G  2.5G   16G  14% /usr/local/nginx/html

###/www 也已经挂载到  /usr/local/nginx/html下

1 2	[iyunv@node2 crmsh-2.3.2]# curl 172.25.0.100 node

###访问web资源也没问题了，说明实现成功

在node1上把pacemaker和corosync重启

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

[iyunv@node1 ~]# crm crm(live)# status Stack: corosync Current DC: node2 (version 1.1.16-12.el7_4.4-94ff4df) - partition with quorum Last updated: Wed Oct 25 21:00:40 2017 Last change: Wed Oct 25 16:56:52 2017 by root via cibadmin on node1 2 nodes configured 3 resources configured Online: [ node1 node2 ] Full list of resources: Resource Group: webservice      webip(ocf::heartbeat:IPaddr):Started node2      webserver(systemd:nginx):Started node2 webstore(ocf::heartbeat:Filesystem):Started node2 crm(live)#

###可以看到node2已经接管了。

四、其他优化

如果设置抢占模式可以这样设

1	crm(live)configure# location nginx_in_node1 nginx inf: node1   ###位置绑定，慎用

服务管理

1	crm(live)configure# property  migration-limit=1      ###当本地服务停掉了，将会本地服务一次，如果起不来就换到另一主机的服务。

crm更改文件

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

crm(live)# configure crm(live)configure# edit    ###会进入配置文件，模式相当于vim的模式 node 1: node1 \         attributes standby=off node 2: node2 primitive webip IPaddr \         params ip=172.25.0.100 primitive webserver systemd:nginx \         op monitor interval=30s timeout=100s \         op start timeout=60s interval=0 \         op stop timeout=60s interval=0 primitive webstore Filesystem \         params device="172.25.0.29:/www" directory="/usr/local/nginx/html" fstype=nfs \         op start timeout=60s interval=0 \         op stop timeout=60s interval=0 \         op monitor interval=20s timeout=40s group webservice webip webserver order webserver_after_webstore Mandatory: webstore webserver colocation webserver_with_webstore_and_webip inf: webserver ( webip webstore ) order webstore_after_webip Mandatory: webip webstore property cib-bootstrap-options: \         have-watchdog=false \         dc-version=1.1.16-12.el7_4.4-94ff4df \         cluster-infrastructure=corosync \         cluster-name=mycluster \         stonith-enabled=false \         migration-limit=1

###可以看到刚刚配的内容  ，可以增删修改。

以上所有是我基于pacemaker+corosync实现nfs+nginx部署内容。