corosync+pacemaker+crmsh+DRBD实现数据库服务器高可用集群构建

y45t4r3

DRBD (DistributedReplicated Block Device) 是 Linux 平台上的分散式储存系统。其中包含了核心模组，数个使用者空间管理程式及 shell scripts，通常用于高可用性（high availability, HA）丛集。DRBD 类似磁盘阵列的RAID 1（镜像），只不过 RAID 1 是在同一台电脑内，而 DRBD 是透过网络。
DRBD 是以 GPL2 授权散布的自由软件。


实验架构图：



一.高可用集群构建的前提条件
1.主机名互相解析，实现主机名通信

1 2 3 4 5

[iyunv@node1 ~]# vim /etc/hosts 127.0.0.1  localhost localhost.localdomain localhost4 localhost4.localdomain4 ::1        localhost localhost.localdomain localhost6 localhost6.localdomain6 172.16.31.10 node1.stu31.com node1 172.16.31.11 node2.stu31.com node2

复制一份到node2：

1	[iyunv@node1 ~]# scp /etc/hostsroot@172.16.31.11:/etc/hosts

2.节点直接实现ssh无密钥通信

1 2 3 4 5 6

节点1： [iyunv@node1 ~]# ssh-keygen -t rsa -P"" [iyunv@node1 ~]# ssh-copy-id -i.ssh/id_rsa.pub root@node2 节点2： [iyunv@node2 ~]# ssh-keygen -t rsa -P"" [iyunv@node2 ~]# ssh-copy-id -i.ssh/id_rsa.pub root@node1

测试ssh无密钥通信：

1 2 3

[iyunv@node2 ~]# date ; ssh node1 'date' Fri Jan 2 12:34:02 CST 2015 Fri Jan 2 12:34:02 CST 2015

时间同步，上面两个节点的时间是一致的！


二.DRBD软件的安装
1.获取DRBD软件程序，CentOS6.6的内核版本是2.6.32-504

1 2	[iyunv@node1 ~]# uname -r 2.6.32-504.el6.x86_64

DRBD已经合并到linux kernel2.6.33及以后内核版本中，这里直接安装管理工具即可，若内核
版本低于2.6.33时请额外安装DRBD内核模块，且与管理工具版本保持一致。

1 2	kmod-drbd84-8.4.5-504.1.el6.x86_64.rpm drbd84-utils-8.9.1-1.el6.elrepo.x86_64.rpm

此软件包是经过编译源码而成，我提供下载，根据附件下载即可：


2.安装软件包，节点1和节点2都需要安装
安装时间将持续很长时间：

1 2 3 4 5 6 7 8 9

[iyunv@node1 ~]# rpm -ivhdrbd84-utils-8.9.1-1.el6.elrepo.x86_64.rpm kmod-drbd84- 8.4.5-504.1.el6.x86_64.rpm warning:drbd84-utils-8.9.1-1.el6.elrepo.x86_64.rpm: Header V4 DSA/SHA1 Signature, key ID baadae52: NOKEY Preparing...               ########################################### [100%]   1:drbd84-utils          ########################################### [ 50%]   2:kmod-drbd84           ########################################### [100%] Working. This may take some time ... Done.

3.各节点准备存储设备
节点1和节点2都需要操作：

1 2 3 4 5 6 7 8

[iyunv@node1 ~]#  echo -n -e "np3+1Gw"|fdisk /dev/sda   [iyunv@node1 ~]# partx -a /dev/sda BLKPG: Device or resource busy error adding partition 1 BLKPG: Device or resource busy error adding partition 2 BLKPG: Device or resource busy error adding partition 3

四.配置DRBD
1.DRBD的配置文件：

1 2 3 4 5

[iyunv@node1 ~]# vim /etc/drbd.conf # You can find an example in  /usr/share/doc/drbd.../drbd.conf.example include "drbd.d/global_common.conf"; include "drbd.d/*.res";

DRBD的所有的控制都是在配置文件/etc/drbd.conf中。通常情况下配置文件包含如下内容：
include"/etc/drbd.d/global_common.conf";
include "/etc/drbd.d/*.res";
通常情况下，/etc/drbd.d/global_common.conf包含global和common的DRBD配置部分，而.res文件都包含一个资源的部分。
在一个单独的drbd.conf文件中配置全部是可以实现的，但是占用的配置很快就会变得混乱，变得难以管理，这也是为什么多文件管理作为首选的原因之一。
无论采用哪种方式，需必须保持在各个集群节点的drbd.conf以及其他的文件完全相同。

2.配置DRBD的全局及通用资源配置文件

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 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74

[iyunv@node1 drbd.d]# cat global_common.conf # DRBD is the result of over a decade ofdevelopment by LINBIT. # In case you need professional servicesfor DRBD or have # feature requests visithttp://www.linbit.com global {          #用于统计应用各个版本的信息。当新的版本的drbd被安装就会和http server进行联系 。当然也可以禁用该选项，默认情况下是启用该选项的。        usage-count no;        # minor-count dialog-refresh disable-ip-verification } common {        handlers {                 # These are EXAMPLE handlersonly.                 # They may have severeimplications,                 # like hard resetting the nodeunder certain circumstances.                 # Be careful when chosing yourpoison.                 #一旦节点发生错误就降级                 pri-on-incon-degr"/usr/lib/drbd/notify-pri-on-incon-degr.sh; /usr/lib/drbd/notify-emergency-reboot.sh;echo b > /proc/sysrq-trigger ; reboot -f";                 #一旦节点发生脑裂的处理是重启                 pri-lost-after-sb"/usr/lib/drbd/notify-pri-lost-after-sb.sh; /usr/lib/drbd/notify-emergency-reboot.sh;echo b > /proc/sysrq-trigger ; reboot -f";                 #一旦本地io错误的处理是关机                 local-io-error"/usr/lib/drbd/notify-io-error.sh; /usr/lib/drbd/notify-emergency-shutdown.sh;echo o > /proc/sysrq-trigger ; halt -f";                 # fence-peer"/usr/lib/drbd/crm-fence-peer.sh";                 # split-brain"/usr/lib/drbd/notify-split-brain.sh root";                 # out-of-sync"/usr/lib/drbd/notify-out-of-sync.sh root";                 # before-resync-target"/usr/lib/drbd/snapshot-resync-target-lvm.sh -p 15 -- -c 16k";                 # after-resync-target/usr/lib/drbd/unsnapshot-resync-target-lvm.sh;        }         startup {                 # wfc-timeout degr-wfc-timeoutoutdated-wfc-timeout wait-after-sb        }        options {                 # cpu-maskon-no-data-accessible        }        disk {                 # size on-io-error fencingdisk-barrier disk-flushes                    #一旦本地磁盘发生IO错误时的操作:拆除                 on-io-error detach;                 # disk-drain md-flushesresync-rate resync-after al-extents                 # c-plan-ahead c-delay-targetc-fill-target c-max-rate                 # c-min-rate disk-timeout        }        net {                 # protocol timeoutmax-epoch-size max-buffers unplug-watermark                 #资源配饰使用完全同步复制协议（Protocol C），除非另有明确指定；表示 收到远程主机的写入确认后,则认为写入完成.                 protocol C;                 # connect-int ping-intsndbuf-size rcvbuf-size ko-count                 # allow-two-primariescram-hmac-alg shared-secret after-sb-0pri                 #设置主备机之间通信使用的信息算法.                 cram-hmac-alg "sha1";                 #消息摘要认证密钥                 shared-secret "password";                 # after-sb-1pri after-sb-2prialways-asbp rr-conflict                 # ping-timeoutdata-integrity-alg tcp-cork on-congestion                 # congestion-fillcongestion-extents csums-alg verify-alg                 # use-rle        }        syncer  {                 #设置主备节点同步时的网络速率最大值,单位是字节.                 rate 1000M;        } }

3.定义节点存储资源配置文件
一个DRBD设备(即:/dev/drbdX),叫做一个"资源"。里面包含一个DRBD设备的主备节点的的ip信息，底层存储设备名称，设备大小，meta信息存放方式，drbd对外提供的设备名等等。

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

[iyunv@node1 drbd.d]# vim mystore.res resource mystore {        #每个主机的说明以"on"开头,后面是主机名.在后面的{}中为这个主机的配置.        on node1.stu31.com  {                 device /dev/drbd0;                 disk /dev/sda3;                 #设置DRBD的监听端口,用于与另一台主机通信                 address 172.16.31.10:7789;                 meta-disk internal;        }        on node2.stu31.com  {                 device /dev/drbd0;                 disk /dev/sda3;                 address 172.16.31.11:7789;                 meta-disk internal;        } }

配置完成后复制一份到节点2：

1 2 3 4 5

[iyunv@node1 drbd.d]# ls global_common.conf  mystore.res [iyunv@node1 drbd.d]# scp *node2:/etc/drbd.d/ global_common.conf                            100% 2105     2.1KB/s  00:00    mystore.res                                   100%  318    0.3KB/s   00:00

4.创建matadata
在启动DRBD之前,需要分别在两台主机的sda分区上,创建供DRBD记录信息的数据块.分别在两台主机上执行:

1 2 3 4 5 6 7 8 9 10 11

[iyunv@node1 drbd.d]# drbdadm create-mdmystore initializing activity log NOT initializing bitmap Writing meta data... New drbd meta data block successfullycreated. [iyunv@node2 ~]# drbdadm create-md mystore initializing activity log NOT initializing bitmap Writing meta data... New drbd meta data block successfullycreated.

5.启动DRBD服务

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

[iyunv@node1 ~]#  /etc/init.d/drbd start Starting DRBD resources: [     create res: mystore   prepare disk: mystore    adjust disk: mystore     adjust net: mystore ] .......... *************************************************************** DRBD's startup script waits for the peernode(s) to appear. - Incase this node was already a degraded cluster before the    rebootthe timeout is 0 seconds. [degr-wfc-timeout] - Ifthe peer was available before the reboot the timeout will   expire after 0 seconds. [wfc-timeout]   (These values are for resource 'mystore'; 0 sec -> wait forever) Toabort waiting enter 'yes' [  21]: . [iyunv@node1 ~]# 节点2启动drbd： [iyunv@node2 ~]# /etc/init.d/drbd start Starting DRBD resources: [     create res: mystore   prepare disk: mystore    adjust disk: mystore     adjust net: mystore ] .

6. 查看DRBD的状态，分别在两台主机上执行

1 2 3 4 5 6

[iyunv@node1 ~]# cat /proc/drbd version: 8.4.5 (api:1/proto:86-101) GIT-hash:1d360bde0e095d495786eaeb2a1ac76888e4db96 build by root@node1.magedu.com, 2015-01-02 12:06:20 0:cs:Connected ro:Secondary/Secondary ds:Inconsistent/Inconsistent C r-----    ns:0 nr:0 dw:0 dr:0 al:0 bm:0 lo:0 pe:0 ua:0 ap:0 ep:1 wo:f oos:1059216

对输出的含义解释如下：
ro表示角色信息，第一次启动drbd时，两个drbd节点默认都处于Secondary状态，
ds是磁盘状态信息，“Inconsistent/Inconsisten”，即为“不一致/不一致”状态，表示两个节点的磁盘数据处于不一致状态。
Ns表示网络发送的数据包信息。
Dw是磁盘写信息
Dr是磁盘读信息

7.设置主节点
由于默认没有主次节点之分，因而需要设置两个主机的主次节点，选择需要设置为主节点的主机，然后执行如下
node1为主节点
#强制设置主节点
[iyunv@node1 ~]# drbdadm primary --forcemystore
查看同步操作：
[iyunv@node1 ~]# drbd-overview
0:mystore/0 SyncSource Primary/Secondary UpToDate/Inconsistent
       [=====>..............] sync'ed: 32.1% (724368/1059216)K
[iyunv@node1 ~]# watch -n1 'cat /proc/drbd'

完成后查看节点状态：

1 2 3 4 5 6

[iyunv@node1 ~]# cat /proc/drbd version: 8.4.5 (api:1/proto:86-101) GIT-hash:1d360bde0e095d495786eaeb2a1ac76888e4db96 build by root@node1.magedu.com, 2015-01-02 12:06:20 0:cs:Connected ro:Primary/Secondary ds:UpToDate/UpToDate C r-----    ns:1059216 nr:0 dw:0 dr:1059912 al:0 bm:0 lo:0 pe:0 ua:0 ap:0 ep:1 wo:foos:0

8.格式化存储

1	[iyunv@node1 ~]# mke2fs -t ext4 /dev/drbd0

1 2 3 4 5 6

挂载到一个目录： [iyunv@node1 ~]# mount /dev/drbd0 /mnt 复制一个文件到mnt： [iyunv@node1 ~]# cp /etc/issue /mnt 卸载存储： [iyunv@node1 ~]# umount /mnt

9.切换主节点为备节点，将node2提升为主节点

1 2 3 4 5 6 7 8

节点1设置为备节点： [iyunv@node1 ~]# drbdadm secondary mystore [iyunv@node1 ~]# drbd-overview 0:mystore/0 Connected Secondary/Secondary UpToDate/UpToDate 提升节点2为主节点： [iyunv@node2 ~]# drbdadm primary mystore [iyunv@node2 ~]# drbd-overview 0:mystore/0 Connected Primary/Secondary UpToDate/UpToDate

挂载文件系统，查看文件是否存在：
[iyunv@node2 ~]# mount /dev/drbd0 /mnt
[iyunv@node2 ~]# ls /mnt
issue lost+found

注意：
（1）mount drbd设备以前必须把设备切换到primary状态。
（2）两个节点中，同一时刻只能有一台处于primary状态，另一台处于secondary状态。
（3）处于secondary状态的服务器上不能加载drbd设备。
（4）主备服务器同步的两个分区大小最好相同,这样不至于浪费磁盘空间,因为drbd磁盘镜像相当于网络raid 1。

10.将drbd服务关闭，开机自启动关闭：

1 2 3 4 5 6 7 8 9

节点1： [iyunv@node1 ~]# service drbd stop Stopping all DRBD resources: . [iyunv@node1 ~]# chkconfig drbd off 节点2： [iyunv@node2 ~]# service drbd stop Stopping all DRBD resources: . [iyunv@node2 ~]# chkconfig drbd off

五.corosync+pacemaker+drbd实现mariadb高可用集群

1.安装corosync和pacemaker软件包:节点1和节点2都安装

1	# yum install corosync pacemaker -y

2.创建配置文件并配置

1 2	[iyunv@node1 ~]# cd /etc/corosync/ [iyunv@node1 corosync]# cpcorosync.conf.example corosync.conf

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

[iyunv@node1 corosync]# cat corosync.conf # Please read the corosync.conf.5 manualpage compatibility: whitetank totem {        version: 2        # secauth: Enable mutual node authentication. If you choose to        # enable this ("on"), then do remember to create a shared        # secret with "corosync-keygen".        #开启认证        secauth: on        threads: 0        # interface: define at least one interface to communicate        # over. If you define more than one interface stanza, you must        # also set rrp_mode.        interface {                 # Rings must be consecutivelynumbered, starting at 0.                 ringnumber: 0                 # This is normally the*network* address of the                 # interface to bind to. Thisensures that you can use                 # identical instances of thisconfiguration file                 # across all your clusternodes, without having to                 # modify this option.                 #定义网络地址                 bindnetaddr: 172.16.31.0                 # However, if you have multiplephysical network                 # interfaces configured for thesame subnet, then the                 # network address alone is notsufficient to identify                 # the interface Corosync shouldbind to. In that case,                 # configure the *host* addressof the interface                 # instead:                 # bindnetaddr: 192.168.1.1                 # When selecting a multicastaddress, consider RFC                 # 2365 (which, among otherthings, specifies that                 # 239.255.x.x addresses areleft to the discretion of                 # the network administrator).Do not reuse multicast                 # addresses across multipleCorosync clusters sharing                 # the same network.                 #定义组播地址                 mcastaddr: 239.224.131.31                 # Corosync uses the port youspecify here for UDP                 # messaging, and also theimmediately preceding                 # port. Thus if you set this to5405, Corosync sends                 # messages over UDP ports 5405and 5404.                 #信息传递端口                 mcastport: 5405                 # Time-to-live for clustercommunication packets. The                 # number of hops (routers) thatthis ring will allow                 # itself to pass. Note thatmulticast routing must be                 # specifically enabled on mostnetwork routers.                 ttl: 1        } } logging {        # Log the source file and line where messages are being        # generated. When in doubt, leave off. Potentially useful for        # debugging.        fileline: off        # Log to standard error. When in doubt, set to no. Useful when        # running in the foreground (when invoking "corosync -f")        to_stderr: no        # Log to a log file. When set to "no", the "logfile"option        # must not be set.        #定义日志记录存放        to_logfile: yes        logfile: /var/log/cluster/corosync.log        # Log to the system log daemon. When in doubt, set to yes.        #to_syslog: yes        # Log debug messages (very verbose). When in doubt, leave off.        debug: off        # Log messages with time stamps. When in doubt, set to on        # (unless you are only logging to syslog, where double        # timestamps can be annoying).        timestamp: on        logger_subsys {                 subsys: AMF                debug: off        } } #以插件方式启动pacemaker： service {        ver:    0        name:   pacemaker }

3.生成认证密钥文件：认证密钥文件需要1024字节，手动写入太麻烦了,我们可以下载程序包来实现写满内存的熵池实现，

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

[iyunv@node1 corosync]# corosync-keygen Corosync Cluster Engine Authentication key generator. Gathering 1024 bits for key from/dev/random. Press keys on your keyboard to generateentropy. Press keys on your keyboard to generateentropy (bits = 128). Press keys on your keyboard to generateentropy (bits = 192). Press keys on your keyboard to generateentropy (bits = 256). Press keys on your keyboard to generateentropy (bits = 320). Press keys on your keyboard to generateentropy (bits = 384). Press keys on your keyboard to generateentropy (bits = 448). Press keys on your keyboard to generateentropy (bits = 512). Press keys on your keyboard to generateentropy (bits = 576). Press keys on your keyboard to generateentropy (bits = 640). Press keys on your keyboard to generateentropy (bits = 704). Press keys on your keyboard to generate entropy(bits = 768). Press keys on your keyboard to generateentropy (bits = 832). Press keys on your keyboard to generateentropy (bits = 896). Press keys on your keyboard to generateentropy (bits = 960). Writing corosync key to/etc/corosync/authkey.

随便下载神马程序都行！


完成后将配置文件及认证密钥复制一份到节点2：

1 2 3

[iyunv@node1 corosync]# scp authkeycorosync.conf node2:/etc/corosync/ authkey                                      100%  128     0.1KB/s  00:00    corosync.conf                                 100% 2724     2.7KB/s  00:00

4.启动corosync服务：

1 2 3 4 5

[iyunv@node1 corosync]# service corosyncstart Starting Corosync Cluster Engine(corosync):               [  OK  ] [iyunv@node2 ~]# service corosync start Starting Corosync Cluster Engine(corosync):               [  OK  ]

5.查看日志：
查看corosync引擎是否正常启动：
节点1的启动日志：

1 2 3 4 5 6

[iyunv@node1 corosync]# grep -e"Corosync Cluster Engine" -e "configuration file" /var/log/cluster/corosync.log Jan 02 14:20:28 corosync [MAIN  ] Corosync Cluster Engine ('1.4.7'): startedand ready to provide service. Jan 02 14:20:28 corosync [MAIN  ] Successfully read main configuration file '/etc/corosync/corosync.conf'.

节点2的启动日志：

1 2 3 4 5 6

[iyunv@node2 ~]# grep -e "CorosyncCluster Engine" -e "configuration file" /var/log/cluster/corosync.log Jan 02 14:20:39 corosync [MAIN  ] Corosync Cluster Engine ('1.4.7'): startedand ready to provide service. Jan 02 14:20:39 corosync [MAIN  ] Successfully read main configuration file '/etc/corosync/corosync.conf'.

查看关键字TOTEM，初始化成员节点通知是否发出:

1 2 3 4 5 6 7 8 9

[iyunv@node1 corosync]# grep"TOTEM" /var/log/cluster/corosync.log Jan 02 14:20:28 corosync [TOTEM ]Initializing transport (UDP/IP Multicast). Jan 02 14:20:28 corosync [TOTEM ]Initializing transmit/receive security: libtomcrypt SOBER128/SHA1HMAC (mode 0). Jan 02 14:20:28 corosync [TOTEM ] Thenetwork interface [172.16.31.10] is now up. Jan 02 14:20:28 corosync [TOTEM ] Aprocessor joined or left the membership and a new membership was formed. Jan 02 14:20:37 corosync [TOTEM ] Aprocessor joined or left the membership and a new membership was formed.

查看监听端口5405是否开启：

1 2 3

[iyunv@node1 ~]# ss -tunl |grep 5405 udp   UNCONN     0      0           172.16.31.10:5405                  *:*    udp   UNCONN     0      0        239.224.131.31:5405                 *:*

查看错误日志：

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

[iyunv@node1 ~]# grep ERROR/var/log/cluster/corosync.log #警告信息：将pacemaker以插件运行的告警，忽略即可 Jan 02 14:20:28 corosync [pcmk  ] ERROR: process_ais_conf: You haveconfigured a cluster using the Pacemaker plugin for Corosync.The plugin is not supported in this environment and will be removed very soon. Jan 02 14:20:28 corosync [pcmk  ] ERROR: process_ais_conf:  Please see Chapter 8 of 'Clusters from Scratch'(http://www.clusterlabs.org/doc) for details on using Pacemaker with CMAN Jan 02 14:20:52 [6260] node1.stu31.com    pengine:  notice: process_pe_message:      Configuration ERRORs found during PE processing.  Please run "crm_verify -L" to identify issues. Jan 02 14:20:52 [6260] node1.stu31.com    pengine:  notice: process_pe_message:      Configuration ERRORs found during PE processing.  Please run "crm_verify -L" to identify issues.

1 2 3 4 5 6 7 8 9

[iyunv@node1 ~]# crm_verify -L -V    #无stonith设备的警告信息，可以忽略               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

六.集群配置工具安装：crmsh软件安装

1.配置yum源：我这里存在一个完整的yum源服务器

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

[iyunv@node1 yum.repos.d]# vimcentos6.6.repo [base] name=CentOS $releasever $basearch on localserver 172.16.0.1 baseurl=http://172.16.0.1/cobbler/ks_mirror/CentOS-6.6-$basearch/ gpgcheck=0 [extra] name=CentOS $releasever $basearch extras baseurl=http://172.16.0.1/centos/$releasever/extras/$basearch/ gpgcheck=0 [epel] name=Fedora EPEL for CentOS$releasever$basearch on local server 172.16.0.1 baseurl=http://172.16.0.1/fedora-epel/$releasever/$basearch/ gpgcheck=0 [corosync2] name=corosync2 baseurl=ftp://172.16.0.1/pub/Sources/6.x86_64/corosync/ gpgcheck=0

复制一份到节点2：

1 2	[iyunv@node1 ~]# scp /etc/yum.repos.d/centos6.6.reponode2:/etc/yum.repos.d/ centos6.6.repo                                100%  521    0.5KB/s   00:00

2.安装crmsh软件，2各节点都安装

1 2 3

# yum install -y crmsh # rpm -qa crmsh crmsh-2.1-1.6.x86_64

3.去除上面的stonith设备警告错误：

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

[iyunv@node1 ~]# crm crm(live)# configure crm(live)configure# propertystonith-enabled=false crm(live)configure# verify #双节点需要仲裁，或者忽略(会造成集群分裂） crm(live)configure# propertyno-quorum-policy=ignore crm(live)configure# verify crm(live)configure# commit crm(live)configure# show node node1.stu31.com node node2.stu31.com property cib-bootstrap-options:        dc-version=1.1.11-97629de        cluster-infrastructure="classic openais (with plugin)"        expected-quorum-votes=2        stonith-enabled=false        no-quorum-policy=ignore

无错误信息输出了：

1 2	[iyunv@node1 ~]# crm_verify -L -V [iyunv@node1 ~]#

七.将DRBD定义为集群服务
1.按照集群服务的要求，首先确保两个节点上的drbd服务已经停止，且不会随系统启动而自动启动：

1 2 3 4

[iyunv@node1 ~]# drbd-overview 0:mystore/0 Unconfigured . . [iyunv@node1 ~]# chkconfig --list drbd drbd           0:off   1:off  2:off   3:off   4:off  5:off   6:off

2.配置drbd为集群资源：
提供drbd的RA目前由OCF归类为linbit，其路径为/usr/lib/ocf/resource.d/linbit/drbd。我们可以使用如下命令来查看此RA及RA的meta信息：

1 2 3 4 5 6 7

[iyunv@node1 ~]# crm ra classes lsb ocf / heartbeat linbit pacemaker service stonith [iyunv@node1 ~]# crm ra list ocf linbit drbd

下面命令可以查看详细信息
[iyunv@node1 ~]# crm ra info ocf:linbit:drbd
输出内容略

drbd需要同时运行在两个节点上，但只能有一个节点（primary/secondary模型）是Master，而另一个节点为Slave；因此，它是一种比较特殊的集群资源，其资源类型为多态（Multi-state）clone类型，即主机节点有Master和Slave之分，且要求服务刚启动时两个节点都处于slave状态。

开始定义集群资源：

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

[iyunv@node1 ~]# crm configure crm(live)configure# primitive mydrbdocf:linbit:drbd params drbd_resource="mystore" op monitor role=Slave interval=20stimeout=20s op monitor role=Master interval=10s timeout=20s op start timeout=240s op stoptimeout=100s crm(live)configure# verify 将集群资源设置为主从模式： crm(live)configure# ms ms_mydrbd mydrbdmeta master-max="1" master-node-max="1" clone-max="2"clone-node-max="1" notify="true" crm(live)configure# verify crm(live)configure# show node node1.stu31.com node node2.stu31.com primitive mydrbd ocf:linbit:drbd        params drbd_resource=mystore        op monitor role=Slave interval=20s timeout=20s        op monitor role=Master interval=10s timeout=20s        op start timeout=240s interval=0        op stop timeout=100s interval=0 ms ms_mydrbd mydrbd        meta master-max=1 master-node-max=1 clone-max=2 clone-node-max=1notify=true property cib-bootstrap-options:        dc-version=1.1.11-97629de        cluster-infrastructure="classic openais (with plugin)"        expected-quorum-votes=2        stonith-enabled=false        no-quorum-policy=ignore crm(live)configure# commit crm(live)configure# cd crm(live)# status Last updated: Sat Jan  3 11:22:54 2015 Last change: Sat Jan  3 11:22:50 2015 Stack: classic openais (with plugin) Current DC: node1.stu31.com - partitionwith quorum Version: 1.1.11-97629de 2 Nodes configured, 2 expected votes 2 Resources configured Online: [ node1.stu31.com node2.stu31.com ] Master/Slave Set: ms_mydrbd [mydrbd]     Masters: [ node2.stu31.com ]     Slaves: [ node1.stu31.com ]

#master-max：有几个主资源master-node-max: 1个节点上最多运行的主资源
#clone-max：有几个克隆资源clone-node-max:1个节点上最多运行的克隆资源
#主从资源也是克隆资源的一种的，只不过它有主从关系

查看drbd的主从状态：

1 2 3 4 5

[iyunv@node1 ~]# drbd-overview 0:mystore/0 Connected Secondary/Primary UpToDate/UpToDate [iyunv@node2 ~]# drbd-overview 0:mystore/0 Connected Primary/Secondary UpToDate/UpToDate

将node2降级成从节点并上线：

1 2 3 4 5 6

[iyunv@node2 ~]# crm node standby [iyunv@node2 ~]# drbd-overview 0:mystore/0 Unconfigured . . [iyunv@node2 ~]# crm node online [iyunv@node2 ~]# drbd-overview 0:mystore/0  Connected Secondary/Primary UpToDate/UpToDate

那么node1就成为主节点了：

1 2	[iyunv@node1 ~]# drbd-overview 0:mystore/0 Connected Primary/Secondary UpToDate/UpToDate

3.定义DRBD存储自动挂载，主节点在哪里，存储就在哪里，需要定义约束

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

crm(live)# configure crm(live)configure# primitive myfsocf:heartbeat:Filesystem params device=/dev/drbd0 directory=/mydata fstype="ext4"op monitor interval=20s timeout=40s op start timeout=60s op stop timeout=60s crm(live)configure# verify #定义协同约束，主节点在哪里启动，存储就跟随主节点 crm(live)configure# colocation myfs_with_ms_mydrbd_masterinf: myfs ms_mydrbd:Master #定义顺序约束，主角色提升完成后才启动存储 crm(live)configure# orderms_mydrbd_master_before_myfs inf: ms_mydrbd:promote myfs:start crm(live)configure# verify crm(live)configure# commit crm(live)configure# cd crm(live)# status Last updated: Sat Jan  3 11:34:23 2015 Last change: Sat Jan  3 11:34:12 2015 Stack: classic openais (with plugin) Current DC: node1.stu31.com - partitionwith quorum Version: 1.1.11-97629de 2 Nodes configured, 2 expected votes 3 Resources configured Online: [ node1.stu31.com node2.stu31.com ] Master/Slave Set: ms_mydrbd [mydrbd]     Masters: [ node1.stu31.com ]     Slaves: [ node2.stu31.com ] myfs  (ocf::heartbeat:Filesystem):   Started node1.stu31.com

可以知道主节点是node1，存储也是挂载在节点1上的。

查看挂载的目录：文件存在，挂载成功

1 2	[iyunv@node1 ~]# ls /mydata issue lost+found

主从资源，文件系统挂载都完成了，下面就开始安装mariadb数据库了！


八.安装mariadb数据库
1.初始化安装mariadb必须在主节点进行：
创建用户mysql管理数据库及配置数据存储目录权限为mysql，两个节点都需要创建用户

1 2	# groupadd -r -g 306 mysql # useradd -r -g 306 -u 306 mysql

获取mariadb的二进制安装包：
mariadb-10.0.10-linux-x86_64.tar.gz

解压至/usr/local目录中：
[iyunv@node1 ~]# tar xfmariadb-10.0.10-linux-x86_64.tar.gz -C /usr/local/

创建软链接：

1 2	[iyunv@node1 ~]# cd /usr/local [iyunv@node1 local]# ln -svmariadb-10.0.10-linux-x86_64/ mysql

在挂载的DRBD存储上创建数据库数据存放目录：
# chown -R mysql:mysql /mydata/

进入安装目录：

1 2 3 4

[iyunv@node1 local]# cd mysql [iyunv@node1 mysql]# pwd /usr/local/mysql [iyunv@node1 mysql]# chown -R root:mysql ./*

初始化安装mariadb：
[iyunv@node1 mysql]#scripts/mysql_install_db --user=mysql --datadir=/mydata/data

安装完成后查看数据存放目录：

1 2 3

[iyunv@node1 mysql]# ls /mydata/data/ aria_log.00000001  ibdata1     ib_logfile1  performance_schema aria_log_control   ib_logfile0 mysql        test

安装成功！

mariadb配置文件的存放，如果我们希望一个节点的配置文件更改后，备节点同步更新，那么配置文件需要存放在drbd存储上是最合适的！

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

[iyunv@node1 mysql]# mkdir /mydata/mysql/ [iyunv@node1 mysql]# chown -R mysql:mysql  /mydata/mysql/ [iyunv@node1 mysql]# cp support-files/my-large.cnf /mydata/mysql/my.cnf [iyunv@node1 mysql]# vim  /mydata/mysql/my.cnf [mysqld] port            = 3306 datadir = /mydata/data socket          = /tmp/mysql.sock skip-external-locking key_buffer_size = 256M max_allowed_packet = 1M table_open_cache = 256 sort_buffer_size = 1M read_buffer_size = 1M read_rnd_buffer_size = 4M myisam_sort_buffer_size = 64M thread_cache_size = 8 query_cache_size= 16M # Try number of CPU's*2 forthread_concurrency thread_concurrency = 8 innodb_file_per_table = on skip_name_resolve = on

在本地创建软链接指向配置文件目录：

1 2	[iyunv@node1 ~]# ln -sv /mydata/mysql/etc/mysql `/etc/mysql' -> `/mydata/mysql'

服务脚本的创建：

1 2	[iyunv@node1 mysql]# cpsupport-files/mysql.server /etc/init.d/mysqld [iyunv@node1 mysql]# chkconfig --add mysqld

启动服务测试：

1 2	[iyunv@node1 mysql]# service mysqld start Starting MySQL.                                           [  OK  ]

登录mysql创建数据库：

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

[iyunv@node1 mysql]#/usr/local/mysql/bin/mysql Welcome to the MariaDB monitor.  Commands end with ; or g. Your MariaDB connection id is 4 Server version: 10.0.10-MariaDB-log MariaDBServer Copyright (c) 2000, 2014, Oracle, SkySQL Aband others. Type 'help;' or 'h' for help. Type 'c' toclear the current input statement. MariaDB [(none)]> create databasetestdb; Query OK, 1 row affected (0.00 sec) MariaDB [(none)]> q Bye

停止mysql服务器：

1 2	[iyunv@node1 mysql]# service mysqld stop Shutting down MySQL..                                      [ OK  ]

2.节点2也要配置mariadb
切换node1为从节点：

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

[iyunv@node1 ~]# crm node standby [iyunv@node1 ~]# crm status Last updated: Sat Jan  3 12:21:38 2015 Last change: Sat Jan  3 12:21:34 2015 Stack: classic openais (with plugin) Current DC: node1.stu31.com - partitionwith quorum Version: 1.1.11-97629de 2 Nodes configured, 2 expected votes 3 Resources configured Node node1.stu31.com: standby Online: [ node2.stu31.com ] Master/Slave Set: ms_mydrbd [mydrbd]     Masters: [ node2.stu31.com ]     Stopped: [ node1.stu31.com ] myfs  (ocf::heartbeat:Filesystem):   Started node2.stu31.com

让node1从节点上线：

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

[iyunv@node1 ~]# crm node online [iyunv@node1 ~]# crm status Last updated: Sat Jan  3 12:21:52 2015 Last change: Sat Jan  3 12:21:48 2015 Stack: classic openais (with plugin) Current DC: node1.stu31.com - partitionwith quorum Version: 1.1.11-97629de 2 Nodes configured, 2 expected votes 3 Resources configured Online: [ node1.stu31.com node2.stu31.com ] Master/Slave Set: ms_mydrbd [mydrbd]     Masters: [ node2.stu31.com ]     Slaves: [ node1.stu31.com ] myfs  (ocf::heartbeat:Filesystem):   Started node2.stu31.com

mariadb程序包解压：

1 2 3 4 5 6

[iyunv@node2 ~]# tar xfmariadb-10.0.10-linux-x86_64.tar.gz -C /usr/local [iyunv@node2 ~]# cd /usr/local [iyunv@node2 local]# ln -sv mariadb-10.0.10-linux-x86_64/mysql `mysql' ->`mariadb-10.0.10-linux-x86_64/' [iyunv@node2 local]# cd mysql [iyunv@node2 mysql]# chown -R root:mysql ./*

不需要初始化安装了！

查看节点2的存储挂载完成与否：

1 2 3 4 5

[iyunv@node2 local]# ls /mydata/data/ aria_log.00000001  ib_logfile1        mysql-bin.index     testdb aria_log_control   multi-master.info  mysql-bin.state ibdata1            mysql              performance_schema ib_logfile0        mysql-bin.000001   test

成功挂载：

只需要服务脚本了：

1 2 3

[iyunv@node2 mysql]# cpsupport-files/mysql.server /etc/init.d/mysqld [iyunv@node2 mysql]# chkconfig --add mysqld [iyunv@node2 mysql]# chkconfig mysqld off

创建软链接将存储的配置文件定位到/etc/下，方便mysql启动：

1 2	[iyunv@node2 ~]# ln -sv /mydata/mysql//etc/mysql `/etc/mysql' -> `/mydata/mysql/'

启动mysqld服务：

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

[iyunv@node2 ~]# service mysqld start Starting MySQL...                                         [  OK  ] [iyunv@node2 ~]# /usr/local/mysql/bin/mysql Welcome to the MariaDB monitor.  Commands end with ; or g. Your MariaDB connection id is 4 Server version: 10.0.10-MariaDB-log MariaDBServer Copyright (c) 2000, 2014, Oracle, SkySQL Aband others. Type 'help;' or 'h' for help. Type 'c' toclear the current input statement. MariaDB [(none)]> show databases; +--------------------+ | Database           | +--------------------+ | information_schema | | mysql              | | performance_schema | | test               | | testdb             | +--------------------+ 5 rows in set (0.04 sec) MariaDB [(none)]> grant all on *.* to'root'@'172.16.%.%' identified by 'oracle'; Query OK, 0 rows affected (0.00 sec) MariaDB [(none)]> flush privileges; Query OK, 0 rows affected (0.00 sec) MariaDB [(none)]> q Bye

可以看出数据库是有testdb的！！！
同步过来了！我们授权一下远程客户端可以登录！
两个节点都安装好了mariadb，


九.定义mariadb数据库集群服务资源

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

[iyunv@node2 ~]# crm crm(live)# configure #定义数据库集群的VIP crm(live)configure# primitive myipocf:heartbeat:IPaddr params ip="172.16.31.166" op monitor interval=10s timeout=20s crm(live)configure# verify #定义数据库集群的服务资源mysqld crm(live)configure# primitive myserverlsb:mysqld op monitor interval=20s timeout=20s crm(live)configure# verify #将资源加入资源组，进行结合资源在一起 crm(live)configure# group myservice myipms_mydrbd:Master myfs myserver ERROR: myservice refers to missing objectms_mydrbd:Master INFO: resource references incolocation:myfs_with_ms_mydrbd_master updated INFO: resource references inorder:ms_mydrbd_master_before_myfs updated #定义资源顺序约束，启动好myfs资源后再启动myserver资源： crm(live)configure# ordermyfs_before_myserver inf: myfs:start myserver:start crm(live)configure# verify #所有都定义完成后就提交！可能mysql服务启动有点慢，等一下即可！ crm(live)configure# commit crm(live)configure# cd crm(live)# status Last updated: Sat Jan  3 13:42:13 2015 Last change: Sat Jan  3 13:41:48 2015 Stack: classic openais (with plugin) Current DC: node1.stu31.com - partitionwith quorum Version: 1.1.11-97629de 2 Nodes configured, 2 expected votes 5 Resources configured Online: [ node1.stu31.com node2.stu31.com ] Master/Slave Set: ms_mydrbd [mydrbd]     Masters: [ node2.stu31.com ]     Slaves: [ node1.stu31.com ] Resource Group: myservice     myip      (ocf::heartbeat:IPaddr):       Started node2.stu31.com     myfs      (ocf::heartbeat:Filesystem):   Started node2.stu31.com     myserver   (lsb:mysqld):   Started node2.stu31.com

启动完成后，我们在远程客户端上连接数据库进行测试：

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

[iyunv@nfs ~]# mysql -h 172.16.31.166 -uroot -poracle Welcome to the MySQL monitor.  Commands end with ; or g. Your MySQL connection id is 4 Server version: 5.5.5-10.0.10-MariaDB-logMariaDB Server Copyright (c) 2000, 2013, Oracle and/or itsaffiliates. All rights reserved. Oracle is a registered trademark of OracleCorporation and/or its affiliates. Other names may be trademarksof their respective owners. Type 'help;' or 'h' for help. Type 'c' toclear the current input statement. mysql> show databases; +--------------------+ | Database           | +--------------------+ | information_schema | | mysql              | | performance_schema | | test               | | testdb             | +--------------------+ 5 rows in set (0.05 sec) mysql> use testdb Database changed mysql> create table t1 (id int); Query OK, 0 rows affected (0.18 sec) mysql> show tables; +------------------+ | Tables_in_testdb | +------------------+ | t1               | +------------------+ 1 row in set (0.01 sec) mysql> q Bye

将节点2切换为备节点，让node1成为主节点：
[iyunv@node2 ~]# crm node standby
输入切换指令后我们监控node1转换成主节点的过程：



查看节点1的集群状态信息：

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

[iyunv@node1 ~]# crm status Last updated: Sat Jan  3 13:59:38 2015 Last change: Sat Jan  3 13:48:49 2015 Stack: classic openais (with plugin) Current DC: node1.stu31.com - partitionwith quorum Version: 1.1.11-97629de 2 Nodes configured, 2 expected votes 5 Resources configured Node node2.stu31.com: standby Online: [ node1.stu31.com ] Master/Slave Set: ms_mydrbd [mydrbd]     Masters: [ node1.stu31.com ]     Stopped: [ node2.stu31.com ] Resource Group: myservice     myip      (ocf::heartbeat:IPaddr):       Started node1.stu31.com     myfs      (ocf::heartbeat:Filesystem):   Started node1.stu31.com     myserver   (lsb:mysqld):   Started node1.stu31.com

再次远程连接数据库测试：

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

[iyunv@nfs ~]# mysql -h 172.16.31.166 -uroot -poracle Welcome to the MySQL monitor.  Commands end with ; or g. Your MySQL connection id is 4 Server version: 5.5.5-10.0.10-MariaDB-logMariaDB Server Copyright (c) 2000, 2013, Oracle and/or itsaffiliates. All rights reserved. Oracle is a registered trademark of OracleCorporation and/or its affiliates. Other names may be trademarksof their respective owners. Type 'help;' or 'h' for help. Type 'c' toclear the current input statement. mysql> use testdb; Reading table information for completion oftable and column names You can turn off this feature to get aquicker startup with -A Database changed mysql> show tables; +------------------+ | Tables_in_testdb | +------------------+ | t1               | +------------------+ 1 row in set (0.00 sec) mysql> q Bye

百度云附件：drbd程序包.rar   


测试成功，同步完成！

至此，corosync+pacemaker+crmsh+DRBD实现数据库服务器高可用性集群的搭建就完成了！！！