DRBD:Distributed Replicated Block Device 分佈式複製塊設備,原理圖如下
DRBD有主雙架構和雙主架構的,當處於主從架構時,這個設備一定只有一個節點是可以讀寫的,另外的節點是不可讀的,連掛載都不可能,只有一個節點是主的,其它節點都是從的。當做爲主主架構時,需要達到幾個條件,1.在高可用集羣中啓用DRBD; 2. 啓用分佈式文件鎖功能,即需要把磁盤格式化爲集羣文件系統(如GFS2,OCFS2等);3. 把DRBD做成資源。
數據的存儲過程:當某個進程存儲數據時在內核中通過DRBD模塊複製一分通過TCP/IP套接字發送到從節點上去,從節點的DRBD在套接字上接收數據並通過DRBD模塊存儲到磁盤,並把結果返回給主節點 。但是內核傳輸的速度很快,而TCP報文的傳輸速度相對來說要慢得多,於是會把數據存入TCP協議棧的發送緩衝區,然後從緩衝區慢慢傳遞到從節點上。這個數據傳輸過程有三個模型:
A : 異步模型,當內核把DRBD模塊中複製的數據已經以報文方式發送到本地的TCP協議棧的緩衝區,至是從節點的數據是否已經傳送到,保存是否完整,無法保證;
B : 半同步模型,報文已經發送到從節點的套接字上,已經接收,併發往內核,但從節點的存儲過程沒有迴應過程。
C:同步模型,報文發往從節點,從節點已經接收,併發往內核,內核已經把數據報文存儲到磁盤,並把結果返回給主節點,這個是默認的數據傳輸模型,可以保證數據的完整性。
安裝環境:
centos6.5,安裝corosync和pacemaker過程請參見上一篇博客,上一個環境安裝的配置接口是crmsh,這裏安裝的是pcs,安裝pcs比較簡單,#yum install pcs –y 即可
安裝完成後集羣狀態如下:
現在看到兩個集羣節點mysql1 mysql2在線,pacemaker是做爲插件在使用的,沒有stonith設備,只有兩個節點,所以quorum也要忽略
#pcs property set stonith-enabled=false
# pcs property set no-quorum-policy=ignore
#pcs config show 查看配置文件
顯示錯誤還沒有集羣配置文件,stonith已經禁用,quorum已經忽略
安裝drbd:
根據官方文檔,drbd在內核2.6.33版本以後就已經裝載進了內核,不需要再額外安裝drbd,只需要安裝drbdadm管理軟件就行,我這裏用的是centos6.5,內核版本是2.6.32,所以還需要安裝
drbd版本有8.1 8.2 8.3 8.4,我這裏用的是已經做好的rpm包進行安裝,下載目錄:ftp://rpmfind.net/linux/atrpms/
把drbd的兩個rpm包放在/root目錄下,然後執行# yum --nogpgcheck localinstall *.rpm
--nogpgcheck 不執行gpg檢測,因爲在centos裏面沒有認證文件
在兩個節點上安裝完成。加載DRBD模塊
[root@mysql1 ~]# modprobe drbd [root@mysql1 ~]# lsmod | grep drbd drbd 325978 0 libcrc32c 1246 1 drbd
創建一個新的磁盤分區做爲DRBD的共享磁盤,我這裏是xen虛擬機添加的一個磁盤/dev/xvdb,創建的磁盤爲/dev/xvdb1
[root@mysql1 drbd.d]# fdisk /dev/xvdb Device contains neither a valid DOS partition table, nor Sun, SGI or OSF disklabel Building a new DOS disklabel with disk identifier 0x218cde3e. Changes will remain in memory only, until you decide to write them. After that, of course, the previous content won't be recoverable. Warning: invalid flag 0x0000 of partition table 4 will be corrected by w(rite) WARNING: DOS-compatible mode is deprecated. It's strongly recommended to switch off the mode (command 'c') and change display units to sectors (command 'u'). Command (m for help): n Command action e extended p primary partition (1-4) p Partition number (1-4): 1 First cylinder (1-2610, default 1): Using default value 1 Last cylinder, +cylinders or +size{K,M,G} (1-2610, default 2610): Using default value 2610 Command (m for help): p Disk /dev/xvdb: 21.5 GB, 21474836480 bytes 255 heads, 63 sectors/track, 2610 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier: 0x218cde3e Device Boot Start End Blocks Id System /dev/xvdb1 1 2610 20964793+ 83 Linux Command (m for help): w The partition table has been altered! Calling ioctl() to re-read partition table. Syncing disks.
在mysql2節點上執行同樣的操作。
配置drbd:
在/etc目錄下有一個文件drbd.conf 主配置文件,裏面主要調用了/etc/drbd.d/目錄下的配置文件
配置/etc/drbd.d/global_common.conf
global { usage-count no; # minor-count dialog-refresh disable-ip-verification } common { protocol C; handlers { 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"; 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"; 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; }#這個配置文件主要是防止腦裂的發生 startup { # wfc-timeout degr-wfc-timeout outdated-wfc-timeout wait-after-sb wfc-timeout 120; degr-wfc-timeout 120; } options { # cpu-mask on-no-data-accessible } disk { on-io-error detach; #fencing resource-only; } net { cram-hmac-alg "sha1"; shared-secret "drbd1"; } syncer { rate 1000M; } }
在/etc/drbd.d/目錄下再創建一個磁盤的資源
#vim /etc/drbd.d/stone.res
resource mysql{ on mysql1 { device /dev/drbd0; disk /dev/xvdb1; address 10.204.80.86:7789; meta-disk internal; } on mysql2 { device /dev/drbd0; disk /dev/xvdb1; address 10.204.80.85:7789; meta-disk internal; } }
把這兩個文件複製到mysql2節點上
然後在兩個節點上初始化資源,分別執行
#drbdadm create-md mysql --這個資源是在stone.res文件中定義的資源
顯示初始化成功
啓動drbd服務
#service drbd start
查看drbd狀態
#cat /proc/drbd
上面的信息兩個節點都處於Secondary 狀態,將mysql1節點設置爲Primary
#drbdadm primary --force mysql
如上所示,數據同步過程已經開始,過一段時間後查看drbd狀態
顯示資源已經是主從狀態
創建文件系統:
#mke2fs –t ext4 –L drbd /dev/drbd0 #格式化爲ext4文件系統,並指定label名爲drbd
創建/data目錄,並掛載分區
#mkdir /data
#mount /dev/drbd0 /data
掛載成功
在/data目錄中創建 a,b,c,d四個文件,然後再手動切換Primary/Secondary節點
在節點mysql1上執行
#cd /data && touch {a,b,c,d} #umount /data #drbdadm secondary mysql
可以看到drbd的狀態又重新變爲secondary /secondary狀態
在mysql2節點上把drbd節點設置爲Primary 狀態,創建/data 目錄 ,並掛載
#drbdadm primary mysql #cat /proc/drbd #mkdir /data #mount /dev/drbd0 /data #ls /data
可以看到在mysql1節點上創建的a,b,c,d四個文件在mysql2節點上依然存在,drbd共享存儲已經Ok
安裝mysql,創建集羣資源,安裝mysql過程略
配置資源資源
pcs創建資源可以用 #pcs ? 可以查看幫助信息
# pcs resource create mysqlip ipaddr ip=10.204.80.89 cidr_netmask=24 nic=eth0
添加DRBD資源,提供drbd的RA目前由OCF歸類爲linbit,其路徑爲/usr/lib/ocf/resource.d/linbit/drbd。drbd 需要同時運行在兩個節點上,但只能有一個節點(primary/secondary模型)是Master,而另一個節點爲Slave;因此,它是一種比較特殊的集羣資源,其資源類型爲多狀態(Multi-state)clone類型,即主機節點有Master和Slave之分,且要求服務剛啓動時兩個節點都處於slave狀態。
#pcs resource create primitive mysqldrbd ocf:linbit:drbd params drbd_resource=mysql op monitor role=Master interval=50s timeout=30s op monitor role=Slave interval=60s timeout=30s op start timeout=240 op stop timeout=100
或者#crm configure primitive mysqldrbd ocf:linbit:drbd params drbd_resource=mysql op monitor role=Master interval=10s timeout=20s op monitor role=Slave interval=20s timeout=20s op start timeout=240 op stop timeout=100
#pcs resource master ms_mysqldrbd mysqldrbd meta master-max="1" master-node-max="1" clone-max="2" clone-node-max="1" notify="true"
或者#crm configure ms ms_myqldrbd mysqldrbd meta master-max="1" master-node-max="1" clone-max="2" clone-node-max="1" notify="true"
把drbd定義爲clone資源,查看資源狀態#pcs resource show
但做爲文件系統還需要掛載,把drbd掛載到/data目錄
# pcs resource create drbdfs ocf:heartbeat:Filesystem params device="/dev/drbd0" directory="/data" fstype="ext4"
查看資源狀態
文件系統掛載drbdfs必須和Master mysqldrbd在同一節點上,必須先啓動mysqldrbd然後才能掛載drbdfs文件系統,所以必須定義資源約束
# pcs constraint colocation add drbdfs with master MS_mysqldrbd
# pcs constraint order promote MS_mysqldrbd then drbdfs ###drbd的資源添加完成
# pcs resource create MySQLD lsb:mysqld op monitor interval=20 timeout=20 on-fail=restart ###添加MySQLD 服務資源完成,定義MySQLD的資源約束,必須drbdfs文件系統先掛載然後才能啓動MySQLD服務
# pcs constraint order drbdfs then MySQLD
把mysqlip ,drbdfs 和MySQLD服務定義到同一個資源組
# pcs resource group add myservice mysqlip drbdfs MySQLD
查看資源狀態
定義資源組的約束
#pcs constraint order mysqlip then promote MS_mysqldrbd then drbdfs then MySQLD
查看集羣的配置信息
#pcs config show
root@mysql1 /]# pcs config show Cluster Name: Corosync Nodes: Pacemaker Nodes: mysql1 mysql2 Resources: Master: ms_mysqldrbd Meta Attrs: master-max=1 master-node-max=1 clone-max=2 clone-node-max=1 notify=true Resource: mysqldrbd (class=ocf provider=linbit type=drbd) Attributes: drbd_resource=mysql Operations: monitor interval=50s role=Master timeout=30s (mysqldrbd-monitor-interval-50s-role-Master) monitor interval=60s role=Slave timeout=30s (mysqldrbd-monitor-interval-60s-role-Slave) Group: myservice Resource: mysqlip (class=ocf provider=heartbeat type=IPaddr) Attributes: ip=10.204.80.89 Operations: monitor interval=60s (mysqlip-monitor-interval-60s) Resource: drbdfs (class=ocf provider=heartbeat type=Filesystem) Attributes: device=/dev/drbd0 directory=/data fstype=ext4 Operations: monitor interval=60s (drbdfs-monitor-interval-60s) Resource: MySQLD (class=lsb type=mysqld) Operations: monitor on-fail=restart interval=20 timeout=20 (MySQLD-monitor-interval-20) Stonith Devices: Fencing Levels: Location Constraints: Ordering Constraints: promote ms_mysqldrbd then start drbdfs (Mandatory) (id:order-ms_mysqldrbd-drbdfs-mandatory) start drbdfs then start MySQLD (Mandatory) (id:order-drbdfs-MySQLD-mandatory) Colocation Constraints: drbdfs with ms_mysqldrbd (INFINITY) (rsc-role:Started) (with-rsc-role:Master) (id:colocation-drbdfs-ms_mysqldrbd-INFINITY-1) Cluster Properties: cluster-infrastructure: classic openais (with plugin) dc-version: 1.1.10-14.el6_5.3-368c726 expected-quorum-votes: 2 no-quorum-policy: ignore stonith-enabled: false
資源狀態情況查看
用其他機器連到10.204.80.89的mysql
已經可以正常連接,把mysql1設置爲standby
發現mysql的集羣已經正常轉移,高可用集羣正常