Heartbeat+drbd+nfs High Availability

First, the environment

System: CentOS 6.4x64 Minimized installation

node1:192.168.1.13

node2:192.168.1.14

vip:192.168.1.15

nfs:192.168.1.10

Second, the basic configuration

Node1 and Node2 do the same thing.

#关闭iptables和selinux [[email protected] ~]# getenforcedisabled                           #确保这项是正确的 [[email protected] ~]# service iptables stop# Configuring local hosts parsing [[email  protected] ~]# echo  "192.168.1.13  node1"  >>/etc/hosts[[email  protected] ~]# echo  "192.168.1.14  node2"  >>/etc/hosts[[email  Protected] ~]# cat /etc/hosts127.0.0.1   localhost localhost.localdomain  localhost4 localhost4.localdomain4::1          localhost localhost.localdomain localhost6 localhost6.localdomain6192.168.1.13   node1192.168.1.14  node2# configuring Epel source [[email protected] ~]# rpm -ivh http:// dl.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm[[email protected] ~]# sed -i  ' [email protected]#[email  Protected]@g '  /etc/yum.repos.d/epel.repo[[email protected] ~]# sed  -i  ' [ Email protected]@#[email protected] '  /etc/yum.repos.d/epel.repo# sync time [[email protected]  ~]# yum install ntp -y[[email protected] ~]# echo  "*/10 *  * * * /usr/sbin/ntpdate asia.pool.ntp.org  &>/dev/null " >/ var/spool/cron/root[[email protected] ~]# ntpdate asia.pool.ntp.org21 jun  17:32:45 ntpdate[1561]: step time server 211.233.40.78 offset -158.552839  sec[[email protected] ~]# hwclock -w# Configuring SSH Trust [[email protected] ~]#  ssh-keygen[[email protected] ~]# ssh-copy-id -i ~/.ssh/id_rsa.pub [email  Protected

Three, installation configuration heartbeat

(1). Install Heartbeat

Perform installation operations #在ha-node1 and Ha-node2 [[email protected] ~]# Yum install heartbeat-y

(2). Configure HA.CF

[[email protected] ~]# cd /usr/share/doc/ heartbeat-3.0.4/[[email protected] heartbeat-3.0.4]# cp authkeys ha.cf  haresources /etc/ha.d/[[email protected] heartbeat-3.0.4]# cd /etc/ha.d/[[email  Protected] ha.d]# lsauthkeys  ha.cf  harc  haresources  rc.d   README.config  resource.d  shellfuncs[[email protected] ha.d]#  egrep -v  "^$|^#"  /etc/ha.d/ha.cf logfile/var/log/ha-loglogfacilitylocal1keepalive  2deadtime 30warntime 10initdead 120mcast eth0 225.0.10.1 694 1 0auto_ Failback onnode node1node node2crm no 

(3). Configure Authkeys

[[email protected] ha.d]# dd if=/dev/random bs=512 count=1 | OpenSSL md50+1 Records in0+1 records out21 bytes (+ B) copied, 3.1278e-05 s, 671 kb/s (stdin) = 4206bd8388c16292bc03710a0c7 47f59[[email protected] ha.d]# grep-v ^#/etc/ha.d/authkeys Auth One MD5 4206bd8388c16292bc03710a0c747f59# Modify the authentication file permissions to 600[ [Email protected] ~]# chmod 600/etc/ha.d/authkeys

(4). Configure Haresource

[Email protected] ha.d]# grep-v ^#/etc/ha.d/haresources node1 Ipaddr::192.168.1.15/24/eth0

(5). Start Heartbeat

[[email protected] ha.d]# scp authkeys haresources  ha.cf node2:/etc/ ha.d/#node1启动服务 [[email protected] ~]# /etc/init.d/heartbeat startstarting  High-availability services: info:  resource is stoppeddone. [[email protected] ~]# chkconfig heartbeat off# Note: Turn off boot, when the server restarts, you need to manually start the #node2 boot service [[ email protected] ~]# /etc/init.d/heartbeat start# view results [[email protected] ~]#  ip a |grep eth02: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu  1500 qdisc pfifo_fast state up qlen 1000    inet  192.168.1.13/24 brd 192.168.1.255 scope global eth0    inet  192.168.1.15/24 brd 192.168.1.255 scope global secondary eth0       #vip在主节点上 [[EMAIL&NBSP;PROTECTED]&NBSP;~]#&NBsp;ip a |grep eth02: eth0: <broadcast,multicast,up,lower_up> mtu 1500  qdisc pfifo_fast state UP qlen 1000    inet  192.168.1.14/24 brd 192.168.1.255 scope global eth0     #备节点上没有vip

(6). Test Heartbeat

Normal state

 #node1信息 [[email protected] ~]# ip a |grep  Eth02: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast  state UP qlen 1000    inet 192.168.1.13/24 brd  192.168.1.255 scope global eth0    inet 192.168.1.15/24 brd  192.168.1.255 scope global secondary eth0      #vip在主节点上 #node2 information [[ Email protected] ~]# ip a |grep eth02: eth0: <broadcast,multicast, up,lower_up> mtu 1500 qdisc pfifo_fast state up qlen 1000     inet 192.168.1.14/24 brd 192.168.1.255 scope global eth0      #备节点上没有vip 

Simulate the status information of the primary node after the outage

#在主节点node1停止heartbeat服务 [[email protected] ~]# /etc/init.d/heartbeat stopstopping  High-availability services: done. [[email protected] ~]# ip a |grep eth0                 #主节点的heartbeat服务停止后, VIP resources were robbed 2: eth0: <broadcast, multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state up qlen 1000     inet 192.168.1.13/24 brd 192.168.1.255 scope global eth0      #在备节点node2查看资源 [[email protected] ~]# ip a |grep eth02:  eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast  state up qlen 1000    inet 192.168.1.14/24 brd 192.168.1.255  scope global eth0    inet 192.168.1.15/24&nbsP;brd 192.168.1.255 scope global secondary eth0 

Heartbeat service to recover the master node

[[email protected] ~]# /etc/init.d/heartbeat startstarting high-availability  Services: info:  resource is stoppeddone. #主节点的heartbeat服务恢复后, took over the resources back [[email  Protected] ~]# ip a |grep eth02: eth0: <broadcast,multicast,up,lower_up > mtu 1500 qdisc pfifo_fast state UP qlen 1000     inet 192.168.1.13/24 brd 192.168.1.255 scope global eth0     inet 192.168.1.15/24 brd 192.168.1.255 scope global secondary eth0      #查看备节点 [[email protected] ~]# ip a |grep eth0           #vip资源已移除2: eth0: <broadcast,multicast,up,lower_up>  mtu 1500 qdisc pfifo_fast state UP qlen 1000     Inet 192.168.1.14/24 brd 192.168.1.255 scope global eth0 

Iv. installation and deployment of DRBD

(1). Partition the hard disk, as Node1 and Node2 do

[[email protected] ~]# fdisk/dev/sdb# Description:/dev/sdb divided into 2 partitions/dev/sdb1 and/dev/sdb2,/dev/sdb1=19g[[email protected] ~]# partprobe/dev/sdb# formatting the partition [[email protected] ~]# MKFS.EXT4/DEV/SDB1 Description: The SDB2 partition is a meta data partition and does not require a format operation [[email protected] ~]# tune2fs-c -1/dev/sdb1 Description: Set maximum mount number to-1, turn off forced check mount count limit

(2). Install DRBD

Since our system is CentOS6.4, we also need to install the kernel module, the version needs to be consistent with UNAME-R, the installation package we extracted from the system installation software, the process is slightly. The installation process of Node1 and Node2 is the same, only the installation process of Node1 is given here.

#安装内核文件 [[email protected] ~]# RPM-IVH kernel-devel-2.6.32-358.el6.x86_64.rpm kernel-headers-2.6.32-358.el6.x86_64. RPM [[email protected] ~]# Yum install drbd84 kmod-drbd84-y

(3). Configuring DRBD

a. Modify the global configuration file

[[email protected] ~]# egrep -v  "^$|^#|^[[:space:]]+#"  /etc/drbd.d/global_ Common.confglobal {usage-count no;} common {protocol c;handlers {}startup {}options {}disk {                 on-io-error detach; no-disk-flushes;no-md-flushes;rate 200m;} net {sndbuf-size 512k;        max-buffers      8000;        unplug-watermark   1024;         max-epoch-size  8000;         cram-hmac-alg  "SHA1";        shared-secret  " weyee2014 ";        after-sb-0pri disconnect;         after-sb-1pri disconnect;        after-sb-2pri disconnect;         rr-conflict disconnect;}}

b. Adding resources

[[email protected] ~]# cat /etc/drbd.d/ nfsdata.resresource nfsdata {    on node1 {         device /dev/drbd1;        disk     /dev/sdb1;        address 192.168.1.13:7789;         meta-disk /dev/sdb2 [0];    }     on node2 {        device /dev/ drbd1;        disk  /dev/sdb1;         address 192.168.1.14:7789;        meta-disk  /dev/sdb2 [0];     }} 

c. Copy the configuration file to Node2, reboot the system to load the DRBD module, initialize the meta data

[[email protected] ~]# scp global_common.conf nfsdata.res node2:/etc/drbd.d/[[ Email protected] ~]# depmod[[email protected] ~]# modprobe drbd[[email  protected] ~]# lsmod |grep drbddrbd                   365931  0 libcrc32c                1246  1 drbd# Initializing meta data in Node1 [[email protected] ~]# drbdadm create-md nfsdatainitializing  Activity lognot initializing bitmapwriting meta data ... new drbd meta data block successfully created. #在node2上加载模块, initialize meta data [[Email  protected] ~]# depmod[[email protected] ~]# modprobe drbd[[email  Protected] ~]# lsmod |grep drbddrbd                  365931   0 libcrc32c                1246  1 drbd[[email protected] ~]# drbdadm create-md  Nfsdatainitializing activity lognot initializing bitmapwriting meta data ... New drbd meta data block successfully created.

D. Start DRBD on Node1 and Node2

#node1操作

This article is from the "ly36843" blog, please be sure to keep this source http://ly36843.blog.51cto.com/3120113/1664020

Heartbeat+drbd+nfs High Availability