全网最新~使用Shell脚本搭建双机高可用的Lustre集群

已于 2023-10-18 12:31:58 修改
阅读量392 收藏 4
点赞数 2
分类专栏： Lustre 文章标签： 服务器 linux 运维 lustre 集群
于 2023-09-26 11:01:33 首次发布
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/qq_56776641/article/details/133300332
版权

1.环境说明：

主机名系统挂载情况网卡ens33，Lnet的IP地址Lustre集群名内存
mds005Centos7.9（共享磁盘）1个mgs，1个MDT，2个OST192.168.10.25/209.21global2G
mds006Centos7.9（共享磁盘）1个mgs，1个MDT，2个OST192.168.10.26/209.22global2G
managerCentos7.9无192.168.10.5无2G

注意：自动化脚本lustre_auto.sh在manager节点上，五块4G共享磁盘（/dev/sdb,/dev/sdc,/dev/sdd,/dev/sde,/dev/sdf），两个网卡（管理IP地址，Lnet通信IP地址）

2.变量定义：

# ens33上的IP地址，即集群管理IP地址
host_address=(192.168.10.25 192.168.10.26)      
# ens38上的IP地址（其他也可以，两个节点网卡名一样就行），即Lnet通信IP地址
lnet_address=(192.168.209.25 192.168.209.26)    
# 集群节点的主机名和域名
host_hostname=(mds005 mds006) 
# ssh连接密码，集群认证密码
host_passwd=110119

3.安装expect命令：

# 安装expect命令
expect -v &> /dev/null
if [ `echo $?` -ne 0 ];then
    echo "没有expect，安装expect命令"
    yum install -y expect
fi

4.配置免密登录：

# 配置免密登录
echo "########################## 本地开始配置ssh ##########################"
if [ `test -a ~/.ssh/id_rsa.pub;echo $?` == 0 ];then
	echo "ssh公钥已创建"
else
	echo "ssh公钥未创建，开始创建"
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10

spawn ssh-keygen -t rsa -b 1024

## 开始进连续捕获
expect	{
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
        "s password:"          { send "${host_passwd}\n"; exp_continue }
        ".ssh/id_rsa)"         { send "\n";  exp_continue }
        "Overwrite (y/n)?"     { send "y\n"; exp_continue }
        "no passphrase):"      { send "\n";  exp_continue }
        "passphrase again:"    { send "\n";  exp_continue }
}
eof
fi



# 本地的密钥开始加入被控制主机
for ((j=0;j<2;j++));do
	echo "########################## ${host_address[j]}正在被添加公钥 ##########################"
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10

spawn ssh-copy-id -i /root/.ssh/id_rsa.pub root@${host_address[j]}

## 开始进连续捕获
expect	{
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
        "s password:"          { send "${host_passwd}\n"; exp_continue }
}
eof
	echo "############# ${host_address[j]}配置完毕 #############"
done




# 被控制主机开始创建密钥
for ((j=0;j<2;j++));do
	echo "########################## ${host_address[j]}开始创建密钥 ##########################"	
	if [ `ssh root@${host_address[j]} 'test -a ~/.ssh/id_rsa.pub;echo $?'` == 0 ];then
		echo "ssh公钥已创建"
	else
		echo "ssh公钥未创建，开始创建"
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10

spawn ssh root@${host_address[j]} "ssh-keygen -t rsa -b 1024"

## 开始进连续捕获
expect	{
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
        "s password:"          { send "${host_passwd}\n"; exp_continue }
        ".ssh/id_rsa)"         { send "\n";  exp_continue }
        "Overwrite (y/n)?"     { send "y\n"; exp_continue }
        "no passphrase):"      { send "\n";  exp_continue }
        "passphrase again:"    { send "\n";  exp_continue }
}
eof
	fi
	echo "############# ${host_address[j]}配置完毕 #############"
done



# 被控制主机开始分配密钥
for ((j=0;j<2;j++));do
	for((k=0;k<2;k++));do
		echo "########################## ${host_address[j]}开始分配公钥给${host_address[k]} ##########################"
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10

spawn ssh -t root@${host_address[j]} "ssh root@${host_address[k]}"
spawn ssh -t root@${host_address[j]} "ssh-copy-id -i /root/.ssh/id_rsa.pub root@${host_address[k]}"

## 开始进连续捕获
expect	{
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
        "s password:"          { send "${host_passwd}\n"; exp_continue }
}
eof
		echo "############# ${host_address[j]}配置完毕 #############"
	done
done

5.主机名域名映射：

for ((i=0;i<100;i++));do
	read -p "修改主机名和配置域名映射？(Y/n): " flag
	if [ "${flag}" == "Y" ];then
		sleep 3
		echo "########################## 开始配置主机名和域名映射 ##########################"
		for ((j=0;j<2;j++));do
			if [ `ssh root@${host_address[j]} "hostname"` != "${host_hostname[j]}" ];then
				ssh root@${host_address[j]} "hostnamectl set-hostname ${host_hostname[j]}"
			fi
			ssh root@${host_address[j]} "cat << eof > /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6
eof"
			for ((k=0;k<2;k++));do
				ssh root@${host_address[j]} "echo '${host_address[k]} ${host_hostname[k]}' >> /etc/hosts"
			done
		done
		echo "############# 配置完毕 #############"
		break
	elif [ "${flag}" == "n" ];then
		echo "############# 已跳过步骤-修改主机名和配置域名映射 #############"
		break
	elif [ ${i} -eq 99 ];then
		echo "############# 已退出 #############"
		exit
	else continue;fi
done

# 测试被控主机ssh连接
for ((j=0;j<2;j++));do
	echo "########################## ${host_address[j]}开始测试 ##########################"	
	for ((k=0;k<2;k++));do
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10

spawn ssh -t root@${host_address[j]} "ssh root@${host_address[k]} 'exit'"

## 开始进连续捕获
expect	{
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
}
eof
		if [ `echo $?` != 0 ];then
			echo "${host_hostname[j]}主机无法免密登录${host_hostname[k]}"
			exit
		fi
	done
	echo "############# ${host_address[j]}测试完毕 #############"
done

6.配置防火墙和selinux：

# 配置防火墙和selinux
for ((i=0;i<100;i++));do
    read -p "配置防火墙和selinux？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置防火墙和selinux ##########################"
        for ((j=0;j<2;j++));do
            ssh root@${host_address[j]} "systemctl stop firewalld;systemctl disable firewalld"
            ssh root@${host_address[j]} "sed -i 's/SELINUX=.*/SELINUX=disabled/' /etc/selinux/config"
        done
        echo "############# 配置完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置防火墙和selinux #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

7.配置yum源：

# 配置yum源
for ((i=0;i<100;i++));do
    read -p "配置yum源？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置yum源 ##########################"
        for ((j=0;j<2;j++));do
            echo "########################## 配置${host_address[j]}的本地yum源 ##########################"
            ssh root@${host_address[j]} "mkdir /mnt/cdrom &> /dev/null;mount /dev/cdrom /mnt/cdrom"
            if [ -z "`ssh root@${host_address[j]} "grep '^\/dev\/cdrom' /etc/fstab"`" ];then
                ssh root@${host_address[j]} "cat << eof >> /etc/fstab
/dev/cdrom /mnt/cdrom iso9660 defaults  0  0
eof"
            fi
            ssh root@${host_address[j]} "cat << eof > /etc/yum.repos.d/centos-local.repo
[centos7.9]
name=centos7.9
baseurl=file:///mnt/cdrom
enabled=1
gpgcheck=0
eof"
            echo "############# ${host_address[j]}配置完毕 #############"
    
            echo "########################## 配置${host_address[j]}的扩展源 ##########################"
            ssh root@${host_address[j]} "yum install epel-release -y"
            echo "############# ${host_address[j]}配置完毕 #############"
    
            echo "########################## 配置${host_address[j]}的阿里yum源 ##########################"
            ssh root@${host_address[j]} "yum install -y wget"
            if [ `ssh root@${host_address[j]} 'test -a /etc/yum.repos.d/CentOS-Base.repo;echo $?'` == 0 ];then
                ssh root@${host_address[j]} "wget -O /etc/yum.repos.d/CentOS-Base.repo http://mirrors.aliyun.com/repo/Centos-7.repo"
            fi
            ssh root@${host_address[j]} "yum clean all && yum repolist"
            echo "############# ${host_address[j]}配置完毕 #############"
    
            if [ `ssh root@${host_address[j]} "echo $?"` != 0 ];then
                echo "yum源配置有误，退出执行脚本"
                exit
            fi
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置yum源 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

8.配置chrony时间服务器：

# 配置chrony时间服务器
for ((i=0;i<100;i++));do
    read -p "配置chrony时间服务器？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置chrony ##########################"
        for ((j=0;j<2;j++));do
            if [ `ssh root@${host_address[j]} "systemctl restart chronyd;echo $?"` != 0 ];then
                echo "${host_address[j]} 安装chrony"
                ssh root@${host_address[j]} "yum install -y chrony && systemctl restart chronyd"
                if [ `echo $?` != 0 ];then
                    echo "安装失败，请排错!"
                    exit
                fi
            fi
            echo "${host_address[j]}配置chrony"
                ssh root@${host_address[j]} "sed -i '/^server/d' /etc/chrony.conf"
            if [ ${host_address[j]} == ${host_address[0]} ];then
                ssh root@${host_address[j]} "sed -i '2a\server '"${host_address[0]}"' iburst\' /etc/chrony.conf"
                ssh root@${host_address[j]} "sed -i 's/#allow 192.168.0.0\/16/allow 192.168.10.0\/16/' /etc/chrony.conf"
                ssh root@${host_address[j]} "sed -i 's/#local stratum 10/local stratum 10/' /etc/chrony.conf"
                sleep 2
​
            else
                ssh root@${host_address[j]} "sed -i '2a\server '"${host_address[0]}"' iburst\' /etc/chrony.conf"
            fi
            ssh root@${host_address[j]} "systemctl restart chronyd && systemctl enable chronyd &> /dev/null"
            sleep 5
            ssh root@${host_address[j]} "timedatectl set-ntp true && chronyc sources -v | sed -n '/^\^\*/p'"
            if [ -z "`ssh root@${host_address[j]} "chronyc sources -v | sed -n '/^\^\*/p'"`" ];then
                echo -e "\e[31m此节点${host_address[j]}的chrony配置有误，请手动调试\e[0m"
                exit
            fi
            echo "############# ${host_address[j]}配置完毕 #############"
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置chrony时间服务器 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

9.安装e2fsprogs：

# 安装e2fsprogs
for ((i=0;i<100;i++));do
    read -p "安装e2fsprogs？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始安装e2fsprogs ##########################"
        for ((j=0;j<2;j++));do
        
            echo "########################## ${host_address[j]}开始安装e2fsprogs ##########################"
            ssh root@${host_address[j]} "rm -rf ~/e2fsprogs1.44.5 && mkdir ~/e2fsprogs1.44.5"
            ssh root@${host_address[j]} "wget -c -r -nd https://downloads.whamcloud.com/public/e2fsprogs/1.44.5.wc1/el7/RPMS/x86_64/ -P ~/e2fsprogs1.44.5"
            ssh root@${host_address[j]} "rm -rf ~/e2fsprogs1.44.5/index.html* ~/e2fsprogs1.44.5/unknown.gif ~/e2fsprogs1.44.5/*.gif ~/e2fsprogs1.44.5/sha256sum"
            ssh root@${host_address[j]} "rpm -Uvh ~/e2fsprogs1.44.5/* --force"
            ssh root@${host_address[j]} "rpm -qa | grep e2fsprogs"
            if [ `echo $?` != 0 ];then
                echo "安装失败，请排错!"
                exit
            fi  
            echo "############# ${host_address[j]}配置完毕 #############"
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-安装e2fsprogs #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

10.安装lustre软件：

for ((i=0;i<100;i++));do
    read -p "安装Lustre软件？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始安装lustre ##########################"
        for ((j=0;j<2;j++));do
            echo "########################## ${host_address[j]}开始安装lustre ##########################"
            ssh root@${host_address[j]} "yum install -y linux-firmware dracut selinux-policy-targeted kexec-tools libyaml perl"
            ssh root@${host_address[j]} "rm -rf ~/lustre2.12.1 && mkdir ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/kernel-3.10.0-957.10.1.el7_lustre.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/kmod-lustre-2.12.1-1.el7.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/kmod-lustre-osd-ldiskfs-2.12.1-1.el7.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/lustre-2.12.1-1.el7.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/lustre-osd-ldiskfs-mount-2.12.1-1.el7.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "rpm -Uvh ~/lustre2.12.1/*.rpm  --force"
            ssh root@${host_address[j]} "rpm -qa | grep lustre"
            if [ `echo $?` != 0 ];then
                echo "安装失败，请排错!"
                exit
            fi
            echo "############# ${host_address[j]}配置完毕 #############"
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-安装lustre #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

11.控制重启和检测重启：

for ((i=0;i<100;i++));do
    read -p "是否重启集群主机（只有重启kernel内核才能更换生效）？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        for((j=0;j<2;j++));do
            ssh root@${host_address[j]} "reboot" 
            continue
        done
    
        for ((k=0;k<100;k++));do
            if [ ${k} -eq 99 ];then
                echo "############# 设备连接超时.... #############"
                exit
            fi
            if [ `ssh root@${host_address[0]} -o ConnectTimeout=5 "exit";echo $?` == 0 -a `ssh root@${host_address[1]} -o ConnectTimeout=5 "exit";echo $?` == 0 ];then
                echo "############# 设备已重启 #############"
                break
            else
                echo "############# 设备正在重启 #############"
            fi
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-重启 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

12.加载模块：

# 检查lustre
for((i=0;i<2;i++));do
    echo "########################## ${host_address[i]}加载Lustre模块，查看Lustre版本 ##########################"
    ssh root@${host_address[i]} "modprobe lustre && lsmod | grep lustre"
    ssh root@${host_address[i]} "modinfo lustre"
    echo "############# ${host_address[i]}配置完毕 #############"
done

13.配置和Lustre网络：

# 配置Lnet网卡的IP地址和Lustre网络
for ((i=0;i<100;i++));do
    read -p "是否继续配置Lnet网卡的IP地址？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        for((j=0;j<2;j++));do
            echo "########################## ${host_address[j]}配置Lnet网卡的IP地址 ##########################"
            read -p "请根据以上输出显示中，输入你要配置的LnNet网卡名称：" network_card
            if [ -z "`ssh root@${host_address[j]} "ip addr | grep -o ${network_card}"`" ];then
                echo "网卡不存在，请重试"
                exit
            fi
            ssh root@${host_address[j]} "nmcli connection delete ${network_card} &> /dev/null"
            ssh root@${host_address[j]} "nmcli connection add type ethernet con-name ${network_card} ifname ${network_card} ipv4.method manual ipv4.addresses '${lnet_address[j]}/24' autoconnect yes"
            ssh root@${host_address[j]} "nmcli connection up ${network_card}"
            echo "############# ${host_address[j]}配置完毕 #############"
            echo "########################## ${host_address[j]}配置Lnet网络 ##########################"
            ssh root@${host_address[j]} "echo options lnet networks='tcp(ens33),tcp2(${network_card})' > /etc/modprobe.d/lustre.conf"
            ssh root@${host_address[j]} "lustre_rmmod && modprobe -v lustre"
            echo "############# ${host_address[j]}配置完毕 #############"
            echo "############# ${host_address[j]}查看Lnet网络 #############"
            ssh root@${host_address[j]} "lctl list_nids"
            echo "##########################################################"
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置Lnet网卡的IP地址 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

14.格式化lustre：

# 格式化lustre
for ((i=0;i<100;i++));do
    read -p "是否继续格式化lustre？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始格式化lustre ##########################"
        ssh root@${host_address[0]} "lsblk"
        ssh root@${host_address[0]} "mkfs.lustre --mgs --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sdb"
        ssh root@${host_address[0]} "mkfs.lustre --fsname global --mdt --index=0 --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --mgsnode=${lnet_address[0]}@tcp2 --mgsnode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sdc"
        ssh root@${host_address[0]} "mkfs.lustre --fsname global --mdt --index=1 --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --mgsnode=${lnet_address[0]}@tcp2 --mgsnode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sdd"
        ssh root@${host_address[0]} "mkfs.lustre --fsname global --ost --index=0 --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --mgsnode=${lnet_address[0]}@tcp2 --mgsnode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sde"
        ssh root@${host_address[0]} "mkfs.lustre --fsname global --ost --index=1 --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --mgsnode=${lnet_address[0]}@tcp2 --mgsnode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sdf"
        echo "############# 格式化完毕 #############"
        echo "########################## 创建挂载点目录 ##########################"
        for((j=0;j<2;j++));do
            ssh root@${host_address[j]} "mkdir /mnt/mgs &> /dev/null;mkdir /mnt/mdt1 &> /dev/null;mkdir /mnt/mdt2 &> /dev/null;mkdir /mnt/ost1 &> /dev/null;mkdir /mnt/ost2 &> /dev/null"
        done
        echo "############# 创建完毕 #############"
        echo "############# ${host_address[1]}查看格式化 #############"
        ssh root@${host_address[1]} "blkid /dev/sdb /dev/sdc /dev/sdd /dev/sde /dev/sdf"
        echo "##########################################################"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-格式化lustre #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

15.测试lustre挂载：

# 测试挂载
for ((i=0;i<100;i++));do
    read -p "是否继续测试lustre的挂载？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        for((j=0;j<2;j++));do
        echo "########################## ${host_address[j]}测试挂载 ##########################"
            echo "########################## mgs测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/mgs &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sdb /mnt/mgs"
            if [ `echo $?` != 0 ];then
                echo "############# mgs测试有误，请手动查看！ #############"
                exit
            fi
            echo "########################## mdt1测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/mdt1 &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sdc /mnt/mdt1"
            if [ `echo $?` != 0 ];then
                echo "############# mdt1测试有误，请手动查看！ #############"
                ssh root@${host_address[j]} "umount /mnt/mgs"
                exit
            fi
            echo "########################## mdt2测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/mdt2 &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sdd /mnt/mdt2"
            if [ `echo $?` != 0 ];then
                echo "############# mdt2测试有误，请手动查看！ #############"
                ssh root@${host_address[j]} "umount /mnt/mdt1;umount /mnt/mgs"
                exit
            fi
            echo "########################## ost1测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/ost1 &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sde /mnt/ost1"
            if [ `echo $?` != 0 ];then
                echo "############# ost1测试有误，请手动查看！ #############"
                ssh root@${host_address[j]} "umount /mnt/mdt2;umount /mnt/mdt1;umount /mnt/mgs"
                exit
            fi
            echo "########################## ost2测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/ost2 &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sdf /mnt/ost2"
            if [ `echo $?` != 0 ];then
                echo "############# ost2测试有误，请手动查看！ #############"
                ssh root@${host_address[j]} "umount /mnt/ost1;umount /mnt/mdt2;umount /mnt/mdt1;umount /mnt/mgs"
                exit
            fi
            ssh root@${host_address[j]} "umount /mnt/ost2;umount /mnt/ost1;umount /mnt/mdt2;umount /mnt/mdt1;umount /mnt/mgs"
        done
        echo "############# 测试完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-测试挂载 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

16.安装高可用和创建集群：

​
# 安装packemaker和corosync软件和创建集群
for ((i=0;i<100;i++));do
    read -p "是否继续安装packemaker和corosync软件和创建集群？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        if [ -z "`ssh root@${host_address[0]} 'pcs status' | grep mycluster`" ];then
            for((j=0;j<2;j++));do
                echo "########################## ${host_address[j]}开始安装 ##########################"
                ssh root@${host_address[j]} "yum install pacemaker pcs policycoreutils-python -y"
                echo "############# ${host_address[j]}安装完毕 #############"
                echo "########################## ${host_address[j]}开始配置 ##########################"
                ssh root@${host_address[j]} "systemctl enable pcsd;systemctl restart pcsd"
                ssh root@${host_address[j]} "echo '${host_passwd}' |passwd --stdin hacluster"
                echo "############# ${host_address[j]}配置完毕 #############"
            done
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10
​
spawn ssh root@${host_address[0]} "pcs cluster auth ${host_hostname[*]}"
​
## 开始进连续捕获
expect  {
        "Username:"          { send "hacluster\n";  exp_continue }
        "Password:"          { send "${host_passwd}\n"; exp_continue }
}
eof
            echo "########################## 开始创建集群 ##########################"
            ssh root@${host_address[0]} "pcs cluster setup --name mylustre ${host_hostname[*]}"
            echo "############# 创建完毕 #############"
        fi
        echo "########################## 启动集群 ##########################"
        ssh root@${host_address[0]} "pcs cluster start --all"
        echo "############# 启动完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-安装packemaker和corosync软件和配置 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

​

17.配置资源防护:

# 配置资源防护
for ((i=0;i<100;i++));do
    read -p "是否继续配置资源防护？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置资源防护 ##########################"
        for((j=0;j<2;j++));do
            ssh root@${host_address[j]} "yum install -y fence-agents-all"
        done
        ssh root@${host_address[0]} "pcs property set stonith-enabled=true"
        if [ `ssh root@${host_address[j]} "pcs status" | grep "stonith:fence_heuristics_ping" | grep -c "Started"` -eq 2 ];then
            echo "############# stonith已创建，并且正常运行，跳过配置stonith #############"
            break
        fi
        if [ `ssh root@${host_address[j]} "pcs status" | grep "stonith:fence_heuristics_ping"` -eq 2 ];then
            ssh root@${host_address[0]} "pcs stonith delete stonith-ping-${host_hostname[0]}"
            ssh root@${host_address[0]} "pcs stonith delete stonith-ping-${host_hostname[0]}"
        fi
​
        ssh root@${host_address[0]} "pcs stonith create stonith-ping-${host_hostname[0]} fence_heuristics_ping ping_targets=${host_address[0]}"
        ssh root@${host_address[0]} "pcs stonith create stonith-ping-${host_hostname[1]} fence_heuristics_ping ping_targets=${host_address[1]}"
        ssh root@${host_address[0]} "pcs status"
        echo "############# 配置完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置资源防护 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

18.创建lustre资源:

# 创建lustre资源
for ((i=0;i<100;i++));do
    read -p "是否继续创建lustre资源（可删除后再创建）？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        for((j=0;j<2;j++));do
            if [ -z `ssh root@${host_address[j]} 'rpm -qa | grep lustre-resource-agents-2.12.1-1'` ];then
                echo "########################## ${host_address[j]}安装ocf:lustre:Lustre包 ##########################"
                ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/lustre-resource-agents-2.12.1-1.el7.x86_64.rpm"
                ssh root@${host_address[j]} "rpm -ivh lustre-resource-agents-2.12.1-1.el7.x86_64.rpm"
                echo "############# 安装完毕 #############"
            fi
        done
        echo "########################## 开始删除lustre资源 ##########################"
        ssh root@${host_address[0]} "pcs resource delete global-ost2 &> /dev/null;pcs resource delete global-ost1 &> /dev/null"
        ssh root@${host_address[0]} "pcs resource delete global-mdt1 &> /dev/null;pcs resource delete global-mdt2 &> /dev/null"
        ssh root@${host_address[0]} "pcs resource delete global-mgs &> /dev/null"
        echo "############# 删除完毕 #############"
        echo "########################## 开始创建lustre资源 ##########################"
        ssh root@${host_address[0]} "pcs resource create global-mgs ocf:lustre:Lustre target=/dev/sdb mountpoint=/mnt/mgs"
        if [ `echo $?` != 0 ];then
            echo "############# mgs资源创建有误，请手动查看！ #############"
            exit
        fi
        ssh root@${host_address[0]} "pcs resource create global-mdt1 ocf:lustre:Lustre target=/dev/sdc mountpoint=/mnt/mdt1"
        if [ `echo $?` != 0 ];then
            echo "############# mdt1资源创建有误，请手动查看！ #############"
            exit
        fi      
        ssh root@${host_address[0]} "pcs resource create global-mdt2 ocf:lustre:Lustre target=/dev/sdd mountpoint=/mnt/mdt2"
        if [ `echo $?` != 0 ];then
            echo "############# mdt2资源创建有误，请手动查看！ #############"
            exit
        fi      
        ssh root@${host_address[0]} "pcs resource create global-ost1 ocf:lustre:Lustre target=/dev/sde mountpoint=/mnt/ost1"
        if [ `echo $?` != 0 ];then
            echo "############# ost1资源创建有误，请手动查看！ #############"
            exit
        fi      
        ssh root@${host_address[0]} "pcs resource create global-ost2 ocf:lustre:Lustre target=/dev/sdf mountpoint=/mnt/ost2"
        if [ `echo $?` != 0 ];then
            echo "############# ost2资源创建有误，请手动查看！ #############"
            exit
        fi      
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-mgs global-mgs ${host_hostname[0]} 10"
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-mdt1 global-mdt1 ${host_hostname[0]} 10"
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-mdt2 global-mdt2 ${host_hostname[1]} 10"
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-ost1 global-ost1 ${host_hostname[0]} 10"
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-ost2 global-ost2 ${host_hostname[1]} 10"
        ssh root@${host_address[0]} "pcs constraint order start global-mgs then start global-mdt1"
        ssh root@${host_address[0]} "pcs constraint order start global-mgs then start global-mdt2"
        ssh root@${host_address[0]} "pcs constraint order start global-mgs then start global-ost1"
        ssh root@${host_address[0]} "pcs constraint order start global-mgs then start global-ost2"
        if [ `echo $?` != 0 ];then
            echo "############# 资源约束创建有误，请手动查看！ #############"
            exit
        fi
        lnet_name1=`ssh root@${host_address[0]} "awk '{print $3}' /etc/modprobe.d/lustre.conf" | awk -F ',' '{print $2}' | awk -F '(' '{print $2}' | awk -F ')' '{print $1}'`
        lnet_name2=`ssh root@${host_address[1]} "awk '{print $3}' /etc/modprobe.d/lustre.conf" | awk -F ',' '{print $2}' | awk -F '(' '{print $2}' | awk -F ')' '{print $1}'`
        ssh root@${host_address[0]} "pcs resource delete ping-lnet &>/dev/null;pcs resource delete global-healthLUSTRE &> /dev/null"
        ssh root@${host_address[0]} "pcs resource create ping-lnet ocf:lustre:healthLNET lctl=true multiplier=1001 device=${lnet_name1} host_list='${lnet_address[0]}@tcp2 ${lnet_address[1]}@tcp2' --clone"
        ssh root@${host_address[0]} "pcs resource create global-healthLUSTRE ocf:lustre:healthLUSTRE --clone"
        if [ `echo $?` != 0 ];then
            echo "############# 资源监听创建有误，请手动查看！ #############"
            exit
        fi
        echo "############# 创建完毕 #############"
        echo "##########################    查看集群状态    ##########################"
        echo "########################################################################"
        echo "########################################################################"
        ssh root@${host_address[0]} "pcs status"
        echo "########################################################################"
        echo "########################################################################"
        echo "################################################## 脚本到此全部执行完毕 ##################################################"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-创建lustre资源 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

19.完整脚本：

# 高可用双机lustre集群的的自动化部署脚本，（部署条件：配置了共享磁盘（/dev/sdb,/dev/sdc,/dev/sdd,/dev/sde,/dev/sdf）。并且有两个网卡，其中一个网卡已经配置了IP地址）
host_address=(192.168.10.25 192.168.10.26)
lnet_address=(192.168.209.25 192.168.209.26)
host_hostname=(mds005 mds006)
host_passwd=110119
​
​
# 安装expect命令
expect -v &> /dev/null
if [ `echo $?` -ne 0 ];then
    echo "没有expect，安装expect命令"
    yum install -y expect
fi
​
# 配置免密登录
echo "########################## 本地开始配置ssh ##########################"
if [ `test -a ~/.ssh/id_rsa.pub;echo $?` == 0 ];then
    echo "ssh公钥已创建"
else
    echo "ssh公钥未创建，开始创建"
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10
​
spawn ssh-keygen -t rsa -b 1024
​
## 开始进连续捕获
expect  {
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
        "s password:"          { send "${host_passwd}\n"; exp_continue }
        ".ssh/id_rsa)"         { send "\n";  exp_continue }
        "Overwrite (y/n)?"     { send "y\n"; exp_continue }
        "no passphrase):"      { send "\n";  exp_continue }
        "passphrase again:"    { send "\n";  exp_continue }
}
eof
fi
​
​
​
# 本地的密钥开始加入被控制主机
for ((j=0;j<2;j++));do
    echo "########################## ${host_address[j]}正在被添加公钥 ##########################"
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10
​
spawn ssh-copy-id -i /root/.ssh/id_rsa.pub root@${host_address[j]}
​
## 开始进连续捕获
expect  {
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
        "s password:"          { send "${host_passwd}\n"; exp_continue }
}
eof
    echo "############# ${host_address[j]}配置完毕 #############"
done
​
​
​
# 被控制主机开始创建密钥
for ((j=0;j<2;j++));do
    echo "########################## ${host_address[j]}开始创建密钥 ##########################"   
    if [ `ssh root@${host_address[j]} 'test -a ~/.ssh/id_rsa.pub;echo $?'` == 0 ];then
        echo "ssh公钥已创建"
    else
        echo "ssh公钥未创建，开始创建"
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10
​
spawn ssh root@${host_address[j]} "ssh-keygen -t rsa -b 1024"
​
## 开始进连续捕获
expect  {
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
        "s password:"          { send "${host_passwd}\n"; exp_continue }
        ".ssh/id_rsa)"         { send "\n";  exp_continue }
        "Overwrite (y/n)?"     { send "y\n"; exp_continue }
        "no passphrase):"      { send "\n";  exp_continue }
        "passphrase again:"    { send "\n";  exp_continue }
}
eof
    fi
    echo "############# ${host_address[j]}配置完毕 #############"
done
​
​
​
# 被控制主机开始分配密钥
for ((j=0;j<2;j++));do
    echo "########################## ${host_address[j]}开始分配公钥 ##########################"   
    for ((k=0;k<2;k++));do
        
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10
​
spawn ssh root@${host_address[j]} "ssh-copy-id -i /root/.ssh/id_rsa.pub root@${host_address[k]}"
​
## 开始进连续捕获
expect  {
        "connecting (yes/no)?" { send "yes\n";  exp_continue }
        "s password:"          { send "${host_passwd}\n"; exp_continue }
}
eof
    done
    echo "############# ${host_address[j]}分配完毕 #############"
done
​
​
​
​
​
# 修改主机名和配置域名映射
for ((i=0;i<100;i++));do
    read -p "修改主机名和配置域名映射？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置主机名和域名映射 ##########################"
        for ((j=0;j<2;j++));do
            if [ `ssh root@${host_address[j]} "hostname"` != "${host_hostname[j]}" ];then
                ssh root@${host_address[j]} "hostnamectl set-hostname ${host_hostname[j]}"
            fi
            ssh root@${host_address[j]} "cat << eof > /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6
eof"
            for ((k=0;k<2;k++));do
                ssh root@${host_address[j]} "echo '${host_address[k]} ${host_hostname[k]}' >> /etc/hosts"
            done
        done
        echo "############# 配置完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-修改主机名和配置域名映射 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
​
​
​
# 配置防火墙和selinux
for ((i=0;i<100;i++));do
    read -p "配置防火墙和selinux？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置防火墙和selinux ##########################"
        for ((j=0;j<2;j++));do
            ssh root@${host_address[j]} "systemctl stop firewalld;systemctl disable firewalld"
            ssh root@${host_address[j]} "sed -i 's/SELINUX=.*/SELINUX=disabled/' /etc/selinux/config"
        done
        echo "############# 配置完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置防火墙和selinux #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
# 配置yum源
for ((i=0;i<100;i++));do
    read -p "配置yum源？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置ssh ##########################"
        for ((j=0;j<2;j++));do
            echo "########################## 配置${host_address[j]}的本地yum源 ##########################"
            ssh root@${host_address[j]} "mkdir /mnt/cdrom &> /dev/null;mount /dev/cdrom /mnt/cdrom"
            if [ -z "`ssh root@${host_address[j]} "grep '^\/dev\/cdrom' /etc/fstab"`" ];then
                ssh root@${host_address[j]} "cat << eof >> /etc/fstab
/dev/cdrom /mnt/cdrom iso9660 defaults  0  0
eof"
            fi
            ssh root@${host_address[j]} "cat << eof > /etc/yum.repos.d/centos-local.repo
[centos7.9]
name=centos7.9
baseurl=file:///mnt/cdrom
enabled=1
gpgcheck=0
eof"
            echo "############# ${host_address[j]}配置完毕 #############"
    
            echo "########################## 配置${host_address[j]}的扩展源 ##########################"
            ssh root@${host_address[j]} "yum install epel-release -y"
            echo "############# ${host_address[j]}配置完毕 #############"
    
            echo "########################## 配置${host_address[j]}的阿里yum源 ##########################"
            ssh root@${host_address[j]} "yum install -y wget"
            if [ `ssh root@${host_address[j]} 'test -a /etc/yum.repos.d/CentOS-Base.repo;echo $?'` == 0 ];then
                ssh root@${host_address[j]} "wget -O /etc/yum.repos.d/CentOS-Base.repo http://mirrors.aliyun.com/repo/Centos-7.repo"
            fi
            ssh root@${host_address[j]} "yum clean all && yum repolist"
            echo "############# ${host_address[j]}配置完毕 #############"
    
            if [ `ssh root@${host_address[j]} "echo $?"` != 0 ];then
                echo "yum源配置有误，退出执行脚本"
                exit
            fi
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置yum源 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
# 配置chrony时间服务器
for ((i=0;i<100;i++));do
    read -p "配置chrony时间服务器？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置chrony ##########################"
        for ((j=0;j<2;j++));do
            if [ `ssh root@${host_address[j]} "systemctl restart chronyd;echo $?"` != 0 ];then
                echo "${host_address[j]} 安装chrony"
                ssh root@${host_address[j]} "yum install -y chrony && systemctl restart chronyd"
                if [ `echo $?` != 0 ];then
                    echo "安装失败，请排错!"
                    exit
                fi
            fi
            echo "${host_address[j]}配置chrony"
                ssh root@${host_address[j]} "sed -i '/^server/d' /etc/chrony.conf"
            if [ ${host_address[j]} == ${host_address[0]} ];then
                ssh root@${host_address[j]} "sed -i '2a\server '"${host_address[0]}"' iburst\' /etc/chrony.conf"
                ssh root@${host_address[j]} "sed -i 's/#allow 192.168.0.0\/16/allow 192.168.10.0\/16/' /etc/chrony.conf"
                ssh root@${host_address[j]} "sed -i 's/#local stratum 10/local stratum 10/' /etc/chrony.conf"
                sleep 2
​
            else
                ssh root@${host_address[j]} "sed -i '2a\server '"${host_address[0]}"' iburst\' /etc/chrony.conf"
            fi
            ssh root@${host_address[j]} "systemctl restart chronyd && systemctl enable chronyd &> /dev/null"
            sleep 5
            ssh root@${host_address[j]} "timedatectl set-ntp true && chronyc sources -v | sed -n '/^\^\*/p'"
            if [ -z "`ssh root@${host_address[j]} "chronyc sources -v | sed -n '/^\^\*/p'"`" ];then
                echo -e "\e[31m此节点${host_address[j]}的chrony配置有误，请手动调试\e[0m"
                exit
            fi
            echo "############# ${host_address[j]}配置完毕 #############"
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置chrony时间服务器 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
​
​
​
# 安装e2fsprogs
for ((i=0;i<100;i++));do
    read -p "安装e2fsprogs？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始安装e2fsprogs ##########################"
        for ((j=0;j<2;j++));do
        
            echo "########################## ${host_address[j]}开始安装e2fsprogs ##########################"
            ssh root@${host_address[j]} "rm -rf ~/e2fsprogs1.44.5 && mkdir ~/e2fsprogs1.44.5"
            ssh root@${host_address[j]} "wget -c -r -nd https://downloads.whamcloud.com/public/e2fsprogs/1.44.5.wc1/el7/RPMS/x86_64/ -P ~/e2fsprogs1.44.5"
            ssh root@${host_address[j]} "rm -rf ~/e2fsprogs1.44.5/index.html* ~/e2fsprogs1.44.5/unknown.gif ~/e2fsprogs1.44.5/*.gif ~/e2fsprogs1.44.5/sha256sum"
            ssh root@${host_address[j]} "rpm -Uvh ~/e2fsprogs1.44.5/* --force"
            ssh root@${host_address[j]} "rpm -qa | grep e2fsprogs"
            if [ `echo $?` != 0 ];then
                echo "安装失败，请排错!"
                exit
            fi  
            echo "############# ${host_address[j]}配置完毕 #############"
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-安装e2fsprogs #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
​
# 安装lustre
for ((i=0;i<100;i++));do
    read -p "安装Lustre软件？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始安装lustre ##########################"
        for ((j=0;j<2;j++));do
            echo "########################## ${host_address[j]}开始安装lustre ##########################"
            ssh root@${host_address[j]} "yum install -y linux-firmware dracut selinux-policy-targeted kexec-tools libyaml perl"
            ssh root@${host_address[j]} "rm -rf ~/lustre2.12.1 && mkdir ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/kernel-3.10.0-957.10.1.el7_lustre.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/kmod-lustre-2.12.1-1.el7.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/kmod-lustre-osd-ldiskfs-2.12.1-1.el7.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/lustre-2.12.1-1.el7.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/lustre-osd-ldiskfs-mount-2.12.1-1.el7.x86_64.rpm -P ~/lustre2.12.1"
            ssh root@${host_address[j]} "rpm -Uvh ~/lustre2.12.1/*.rpm  --force"
            ssh root@${host_address[j]} "rpm -qa | grep lustre"
            if [ `echo $?` != 0 ];then
                echo "安装失败，请排错!"
                exit
            fi
            echo "############# ${host_address[j]}配置完毕 #############"
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-安装lustre #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
#echo "########################## 请手动重启 ##########################"
for ((i=0;i<100;i++));do
    read -p "是否重启集群主机（只有重启kernel内核才能更换生效）？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        for((j=0;j<2;j++));do
            ssh root@${host_address[j]} "reboot" 
            continue
        done
    
        for ((k=0;k<100;k++));do
            if [ ${k} -eq 99 ];then
                echo "############# 设备连接超时.... #############"
                exit
            fi
            if [ `ssh root@${host_address[0]} -o ConnectTimeout=5 "exit";echo $?` == 0 -a `ssh root@${host_address[1]} -o ConnectTimeout=5 "exit";echo $?` == 0 ];then
                echo "############# 设备已重启 #############"
                break
            else
                echo "############# 设备正在重启 #############"
            fi
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-重启 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
sleep 5
​
​
# 检查lustre
for((i=0;i<2;i++));do
    echo "########################## ${host_address[i]}加载Lustre模块，查看Lustre版本 ##########################"
    ssh root@${host_address[i]} "modprobe lustre && lsmod | grep lustre"
    ssh root@${host_address[i]} "modinfo lustre"
    echo "############# ${host_address[i]}配置完毕 #############"
done
​
​
​
# 配置配置Lnet网卡的IP地址和Lustre网络
for ((i=0;i<100;i++));do
    read -p "是否继续配置Lnet网卡的IP地址？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        for((j=0;j<2;j++));do
            echo "########################## ${host_address[j]}配置Lnet网卡的IP地址 ##########################"
            read -p "请根据以上输出显示中，输入你要配置的LnNet网卡名称：" network_card
            if [ -z "`ssh root@${host_address[j]} "ip addr | grep -o ${network_card}"`" ];then
                echo "网卡不存在，请重试"
                exit
            fi
            ssh root@${host_address[j]} "nmcli connection delete ${network_card} &> /dev/null"
            ssh root@${host_address[j]} "nmcli connection add type ethernet con-name ${network_card} ifname ${network_card} ipv4.method manual ipv4.addresses '${lnet_address[j]}/24' autoconnect yes"
            ssh root@${host_address[j]} "nmcli connection up ${network_card}"
            echo "############# ${host_address[j]}配置完毕 #############"
            echo "########################## ${host_address[j]}配置Lnet网络 ##########################"
            ssh root@${host_address[j]} "echo options lnet networks='tcp(ens33),tcp2(${network_card})' > /etc/modprobe.d/lustre.conf"
            ssh root@${host_address[j]} "lustre_rmmod && modprobe -v lustre"
            echo "############# ${host_address[j]}配置完毕 #############"
            echo "############# ${host_address[j]}查看Lnet网络 #############"
            ssh root@${host_address[j]} "lctl list_nids"
            echo "##########################################################"
        done
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置Lnet网卡的IP地址 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
​
​
# 格式化lustre
for ((i=0;i<100;i++));do
    read -p "是否继续格式化lustre？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始格式化lustre ##########################"
        ssh root@${host_address[0]} "lsblk"
        ssh root@${host_address[0]} "mkfs.lustre --mgs --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sdb"
        ssh root@${host_address[0]} "mkfs.lustre --fsname global --mdt --index=0 --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --mgsnode=${lnet_address[0]}@tcp2 --mgsnode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sdc"
        ssh root@${host_address[0]} "mkfs.lustre --fsname global --mdt --index=1 --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --mgsnode=${lnet_address[0]}@tcp2 --mgsnode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sdd"
        ssh root@${host_address[0]} "mkfs.lustre --fsname global --ost --index=0 --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --mgsnode=${lnet_address[0]}@tcp2 --mgsnode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sde"
        ssh root@${host_address[0]} "mkfs.lustre --fsname global --ost --index=1 --servicenode=${lnet_address[0]}@tcp2 --servicenode=${lnet_address[1]}@tcp2 --mgsnode=${lnet_address[0]}@tcp2 --mgsnode=${lnet_address[1]}@tcp2 --backfstype=ldiskfs --reformat /dev/sdf"
        echo "############# 格式化完毕 #############"
        echo "########################## 创建挂载点目录 ##########################"
        for((j=0;j<2;j++));do
            ssh root@${host_address[j]} "mkdir /mnt/mgs &> /dev/null;mkdir /mnt/mdt1 &> /dev/null;mkdir /mnt/mdt2 &> /dev/null;mkdir /mnt/ost1 &> /dev/null;mkdir /mnt/ost2 &> /dev/null"
        done
        echo "############# 创建完毕 #############"
        echo "############# ${host_address[1]}查看格式化 #############"
        ssh root@${host_address[1]} "blkid /dev/sdb /dev/sdc /dev/sdd /dev/sde /dev/sdf"
        echo "##########################################################"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-格式化lustre #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
# 测试挂载
for ((i=0;i<100;i++));do
    read -p "是否继续测试lustre的挂载？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        for((j=0;j<2;j++));do
        echo "########################## ${host_address[j]}测试挂载 ##########################"
            echo "########################## mgs测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/mgs &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sdb /mnt/mgs"
            if [ `echo $?` != 0 ];then
                echo "############# mgs测试有误，请手动查看！ #############"
                exit
            fi
            echo "########################## mdt1测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/mdt1 &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sdc /mnt/mdt1"
            if [ `echo $?` != 0 ];then
                echo "############# mdt1测试有误，请手动查看！ #############"
                ssh root@${host_address[j]} "umount /mnt/mgs"
                exit
            fi
            echo "########################## mdt2测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/mdt2 &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sdd /mnt/mdt2"
            if [ `echo $?` != 0 ];then
                echo "############# mdt2测试有误，请手动查看！ #############"
                ssh root@${host_address[j]} "umount /mnt/mdt1;umount /mnt/mgs"
                exit
            fi
            echo "########################## ost1测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/ost1 &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sde /mnt/ost1"
            if [ `echo $?` != 0 ];then
                echo "############# ost1测试有误，请手动查看！ #############"
                ssh root@${host_address[j]} "umount /mnt/mdt2;umount /mnt/mdt1;umount /mnt/mgs"
                exit
            fi
            echo "########################## ost2测试挂载 ##########################"
            ssh root@${host_address[j]} "umount /mnt/ost2 &> /dev/null"
            ssh root@${host_address[j]} "mount -t lustre /dev/sdf /mnt/ost2"
            if [ `echo $?` != 0 ];then
                echo "############# ost2测试有误，请手动查看！ #############"
                ssh root@${host_address[j]} "umount /mnt/ost1;umount /mnt/mdt2;umount /mnt/mdt1;umount /mnt/mgs"
                exit
            fi
            ssh root@${host_address[j]} "umount /mnt/ost2;umount /mnt/ost1;umount /mnt/mdt2;umount /mnt/mdt1;umount /mnt/mgs"
        done
        echo "############# 测试完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-测试挂载 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
​
# 安装packemaker和corosync软件和创建集群
for ((i=0;i<100;i++));do
    read -p "是否继续安装packemaker和corosync软件和创建集群？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        if [ -z "`ssh root@${host_address[0]} 'pcs status' | grep mycluster`" ];then
            for((j=0;j<2;j++));do
                echo "########################## ${host_address[j]}开始安装 ##########################"
                ssh root@${host_address[j]} "yum install pacemaker pcs policycoreutils-python -y"
                echo "############# ${host_address[j]}安装完毕 #############"
                echo "########################## ${host_address[j]}开始配置 ##########################"
                ssh root@${host_address[j]} "systemctl enable pcsd;systemctl restart pcsd"
                ssh root@${host_address[j]} "echo '${host_passwd}' |passwd --stdin hacluster"
                echo "############# ${host_address[j]}配置完毕 #############"
            done
/usr/bin/expect << eof
# 设置捕获字符串后，期待回复的超时时间
set timeout 10
​
spawn ssh root@${host_address[0]} "pcs cluster auth ${host_hostname[*]}"
​
## 开始进连续捕获
expect  {
        "Username:"          { send "hacluster\n";  exp_continue }
        "Password:"          { send "${host_passwd}\n"; exp_continue }
}
eof
            echo "########################## 开始创建集群 ##########################"
            ssh root@${host_address[0]} "pcs cluster setup --name mylustre ${host_hostname[*]}"
            echo "############# 创建完毕 #############"
        fi
        echo "########################## 启动集群 ##########################"
        ssh root@${host_address[0]} "pcs cluster start --all"
        echo "############# 启动完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-安装packemaker和corosync软件和配置 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
​
# 配置资源防护
for ((i=0;i<100;i++));do
    read -p "是否继续配置资源防护？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        echo "########################## 开始配置资源防护 ##########################"
        for((j=0;j<2;j++));do
            ssh root@${host_address[j]} "yum install -y fence-agents-all"
        done
        ssh root@${host_address[0]} "pcs property set stonith-enabled=true"
        if [ `ssh root@${host_address[j]} "pcs status" | grep "stonith:fence_heuristics_ping" | grep -c "Started"` -eq 2 ];then
            echo "############# stonith已创建，并且正常运行，跳过配置stonith #############"
            break
        fi
        if [ `ssh root@${host_address[j]} "pcs status" | grep "stonith:fence_heuristics_ping"` -eq 2 ];then
            ssh root@${host_address[0]} "pcs stonith delete stonith-ping-${host_hostname[0]}"
            ssh root@${host_address[0]} "pcs stonith delete stonith-ping-${host_hostname[0]}"
        fi
​
        ssh root@${host_address[0]} "pcs stonith create stonith-ping-${host_hostname[0]} fence_heuristics_ping ping_targets=${host_address[0]}"
        ssh root@${host_address[0]} "pcs stonith create stonith-ping-${host_hostname[1]} fence_heuristics_ping ping_targets=${host_address[1]}"
        ssh root@${host_address[0]} "pcs status"
        echo "############# 配置完毕 #############"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-配置资源防护 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done
​
​
# 创建lustre资源
for ((i=0;i<100;i++));do
    read -p "是否继续创建lustre资源（可删除后再创建）？(Y/n): " flag
    if [ "${flag}" == "Y" ];then
        sleep 3
        for((j=0;j<2;j++));do
            if [ -z `ssh root@${host_address[j]} 'rpm -qa | grep lustre-resource-agents-2.12.1-1'` ];then
                echo "########################## ${host_address[j]}安装ocf:lustre:Lustre包 ##########################"
                ssh root@${host_address[j]} "wget https://downloads.whamcloud.com/public/lustre/lustre-2.12.1/el7/server/RPMS/x86_64/lustre-resource-agents-2.12.1-1.el7.x86_64.rpm"
                ssh root@${host_address[j]} "rpm -ivh lustre-resource-agents-2.12.1-1.el7.x86_64.rpm"
                echo "############# 安装完毕 #############"
            fi
        done
        echo "########################## 开始删除lustre资源 ##########################"
        ssh root@${host_address[0]} "pcs resource delete global-ost2 &> /dev/null;pcs resource delete global-ost1 &> /dev/null"
        ssh root@${host_address[0]} "pcs resource delete global-mdt1 &> /dev/null;pcs resource delete global-mdt2 &> /dev/null"
        ssh root@${host_address[0]} "pcs resource delete global-mgs &> /dev/null"
        echo "############# 删除完毕 #############"
        echo "########################## 开始创建lustre资源 ##########################"
        ssh root@${host_address[0]} "pcs resource create global-mgs ocf:lustre:Lustre target=/dev/sdb mountpoint=/mnt/mgs"
        if [ `echo $?` != 0 ];then
            echo "############# mgs资源创建有误，请手动查看！ #############"
            exit
        fi
        ssh root@${host_address[0]} "pcs resource create global-mdt1 ocf:lustre:Lustre target=/dev/sdc mountpoint=/mnt/mdt1"
        if [ `echo $?` != 0 ];then
            echo "############# mdt1资源创建有误，请手动查看！ #############"
            exit
        fi      
        ssh root@${host_address[0]} "pcs resource create global-mdt2 ocf:lustre:Lustre target=/dev/sdd mountpoint=/mnt/mdt2"
        if [ `echo $?` != 0 ];then
            echo "############# mdt2资源创建有误，请手动查看！ #############"
            exit
        fi      
        ssh root@${host_address[0]} "pcs resource create global-ost1 ocf:lustre:Lustre target=/dev/sde mountpoint=/mnt/ost1"
        if [ `echo $?` != 0 ];then
            echo "############# ost1资源创建有误，请手动查看！ #############"
            exit
        fi      
        ssh root@${host_address[0]} "pcs resource create global-ost2 ocf:lustre:Lustre target=/dev/sdf mountpoint=/mnt/ost2"
        if [ `echo $?` != 0 ];then
            echo "############# ost2资源创建有误，请手动查看！ #############"
            exit
        fi      
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-mgs global-mgs ${host_hostname[0]} 10"
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-mdt1 global-mdt1 ${host_hostname[0]} 10"
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-mdt2 global-mdt2 ${host_hostname[1]} 10"
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-ost1 global-ost1 ${host_hostname[0]} 10"
        ssh root@${host_address[0]} "pcs constraint location add global-constraint-ost2 global-ost2 ${host_hostname[1]} 10"
        ssh root@${host_address[0]} "pcs constraint order start global-mgs then start global-mdt1"
        ssh root@${host_address[0]} "pcs constraint order start global-mgs then start global-mdt2"
        ssh root@${host_address[0]} "pcs constraint order start global-mgs then start global-ost1"
        ssh root@${host_address[0]} "pcs constraint order start global-mgs then start global-ost2"
        if [ `echo $?` != 0 ];then
            echo "############# 资源约束创建有误，请手动查看！ #############"
            exit
        fi
        lnet_name1=`ssh root@${host_address[0]} "awk '{print $3}' /etc/modprobe.d/lustre.conf" | awk -F ',' '{print $2}' | awk -F '(' '{print $2}' | awk -F ')' '{print $1}'`
        lnet_name2=`ssh root@${host_address[1]} "awk '{print $3}' /etc/modprobe.d/lustre.conf" | awk -F ',' '{print $2}' | awk -F '(' '{print $2}' | awk -F ')' '{print $1}'`
        ssh root@${host_address[0]} "pcs resource delete ping-lnet &>/dev/null;pcs resource delete global-healthLUSTRE &> /dev/null"
        ssh root@${host_address[0]} "pcs resource create ping-lnet ocf:lustre:healthLNET lctl=true multiplier=1001 device=${lnet_name1} host_list='${lnet_address[0]}@tcp2 ${lnet_address[1]}@tcp2' --clone"
        ssh root@${host_address[0]} "pcs resource create global-healthLUSTRE ocf:lustre:healthLUSTRE --clone"
        if [ `echo $?` != 0 ];then
            echo "############# 资源监听创建有误，请手动查看！ #############"
            exit
        fi
        echo "############# 创建完毕 #############"
        echo "##########################    查看集群状态    ##########################"
        echo "########################################################################"
        echo "########################################################################"
        ssh root@${host_address[0]} "pcs status"
        echo "########################################################################"
        echo "########################################################################"
        echo "################################################## 脚本到此全部执行完毕 ##################################################"
        break
    elif [ "${flag}" == "n" ];then
        echo "############# 已跳过步骤-创建lustre资源 #############"
        break
    elif [ ${i} -eq 99 ];then
        echo "############# 已退出 #############"
        exit
    else continue;fi
done

碧蓝幻想
关注
  • 2
    点赞
  • 踩
  • 4
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
专栏目录
keepalive+nginx+redis实现双机热备加redis主从高可用
qq_36797902的博客
09-07 391
script "/usr/local/keepalived/scripts/redis/redis_check.sh winning.2019 " ###监控脚本。script "/usr/local/keepalived/scripts/redis/redis_check.sh winning.2019 " ###监控脚本。输入命令：vim/winning/winmid/redis/conf/common.conf。三、建立文件夹：/usr/local/keepalived/scripts。
【Lustre相关】应用部署-03-Lustre集群部署实践（软raid方案）
最新发布
Luxf0的博客
11-27 1884
以实际项目作为案例，阐述lustre软raid集群部署过程，包含高可用配置方案介绍及常见配置文件解决方法
1-6改变退出状态码的exit命令
Xx13624558575的博客
10-30 488
书接上文，我们来说一下改变退出状态码的exit命令。我们在上一个小节中给大家讲了，退出状态码通常用于脚本中在if中做判断，便于进一步的逻辑处理。对于退出状态码它有个默认的准则。 退出状态码是以上一条指令的返回结果为准 也就是说上一条指令的执行状况会直接决定我们的退出状态码的值。其实我们可以使用exit命令来改变退出状态码的值。 exit $exit_code 例如：exit 48 或者 exit 125 或者 exit 0 在某些场景下我...
cp linux 显示进度条_cp文件显示进度条
weixin_42350606的博客
12-31 235
#!/bin/bashtrap 'exit_fun' 2usage(){echo "Usage: `basename $0` \"src file\" \"dst file\""exit 1}exit_fun(){echo -e "\033[?25h"kill -9 $(ps -ef|awk '/ [c]p /{print $2}') &>/dev/nullexit 1}[ "$#"...
Linux常用命令——exit命令
AI的博客
01-08 3096
同于退出shell，并返回给定值。在shell脚本中可以终止当前脚本执行。执行exit可使shell以指定的状态值退出。若不设置状态值参数，则shell以预设值退出。状态值0代表执行成功，其他值代表执行失败。返回值：指定shell返回值。退出当前的shell。
exit &amp; $?问题
snow365的专栏
06-12 843
1.如果程序正常结束，会默认返回0，如果有err, 会返回1；当然也可以自己定义返回的数字，供自己使用2.若程序结束要让自动退出窗口， 就可以写明 "exit 0" 或"exit 1";如果只是让程序结束，不用退出窗口，只用"exit" 即可。3.$?：是指最近的命令的执行结果，自己的犯的小错误：   test.sh： ./A.shif [ $? -eq 0 ]     then         
双机高可用Lustre自动化部署脚本（虚拟机，没有IB网络哦~）
09-26
此脚本可以选择性执行各个模块（主机名域名映射，配置防火墙和selinux，配置yum源，配置chrony时间服务器，安装e2fsprogs，安装lustre软件，控制重启和检测重启，配置和Lustre网络，格式化lustre，测试lustre挂载，...
kafka+zookeeper高可用集群搭建shell使用脚本
02-22
【标题】"kafka+zookeeper高可用集群搭建shell使用脚本"所涉及的知识点主要集中在构建高可用的Kafka和Zookeeper集群上，以及如何利用Shell脚本来自动化这个过程。Kafka是一个分布式流处理平台，而Zookeeper是Apache...
一键shell脚本搭建docker redis 集群（cluster）
12-30
为了新手小伙伴发愁集群创建难，特意写了一个shell脚本。自己技术也很菜，但是我坚信也会有跟我一样菜的小伙伴。。。。哈哈哈 不多说直接上脚本 希望有大神更加优化一下，优化后的版本可以分享下。哇 等着你大神，...
nginx实现请求的负载均衡 + keepalived实现nginx的高可用的shell脚本
02-22
运行这个脚本可以在Linux环境下快速部署这样一个高可用的负载均衡集群。 总结起来，Nginx和Keepalived的组合为Web服务提供了强大的性能优化和故障恢复能力。理解并熟练运用这些技术对于提升系统稳定性和用户体验至...
lustre中文手册
12-08
lustre中文手册
用shell脚本实现自动切换内网和外网实现高可用
01-10
首先说明下我们的配置文件，都是类似格式的，假如内网是192.168.0.3，外网是123.123.123.123，配置文件如下： $db[‘salver’][‘hostname’] = ‘192.168.0.3:3306’;//$db[‘salver’][‘hostname’] = ‘123.123....
expect实现scp远程复制文件 spawn
abc
06-02 746
expect实现scp远程复制文件 spawnExpect是一个用来实现自动交互功能的软件套件。使用Expect可以模拟手工交互的过程，实现自动与远端程序的交互。例如当执行ssh命令连接服务器时，需要手动输入密码，如果采用Expect，就可以自动交互，无需再人工手动输入密码了。
Redis双机热备
宿久
02-04 6183
      双机热备特指基于高可用系统中的两台服务器的热备（或高可用），因两机高可用在国内使用较多，故得名双机热备，双机高可用按工作中的切换方式分为：主-备方式（Active-Standby方式）和双主机方式（Active-Active方式），主-备方式即指的是一台服务器处于某种业务的激活状态（即Active状态），另一台服务器处于该业务的备用状态（即Standby状态)。而双主机方式即指两种不同...
redis 4.0.9 centos7 双机集群安装
草宝虫的博客
03-04 547
开启limits限制 sudo bash -c 'cat >> /etc/security/limits.conf <<-EOF * soft nofile 65536 * hard nofile 65536 * soft nproc 131072 * hard nproc 131072 EOF' reboot重启 安装redis 两台机器都运行 yum install -y wget gcc wget http://download.redis.io/releases/redi
Redis(带认证)+keepalived双机热备部署方案
qq_36811981的博客
09-10 2419
1. 前言 目前Redis主流的高可用方案主要是Master+Slave+Sentinel（即主从+哨兵）的模式，主要原理是：通过Sentinel（哨兵）集群来检测集群各节点Redis服务的存活状态及完成Redis集群的主从切换。切换完成后，Sentinel调用notification-script参数指定的配置文件，通知DNS服务更改DNS配置，完成Redis master节点的切换。 然而我们现在的应用不支持Sentinel模式，并且代码改起来改动较大，风险也比较大。所以只好舍弃了Master+Slav
Redis双机主从部署(持久化)及备份迁移
qq_42063179的博客
11-13 623
部署环境 centos7.1虚拟机两台 操作步骤 1.安装需要的组件 yum install -y tcl gcc 2.下载redis的压缩包 wget http://download.redis.io/releases/redis-4.0.9.tar.gz 3.解压到指定目录 tar -zxvfredis-4.0.9.tar.gz -C /data 4.安装 cd /data/redis-4.0.9/src make && make install ti...
lustre的安装
sqlora的专栏
05-07 2260
项目背景介绍 本次项目目的在于构建一个具有高性能、支持高并发读写、文件共享的存储系统。Lustre在HPC领域被广泛使用，本次项目建设也是在调研了其他诸多文件系统之后，最终选择Lustre文件系统作为软件部分，由于Lustre文件系统本身无数据安全机制，必须构建于稳定的磁盘阵列之上，通过硬件的稳定性来解决数据的安全性。 硬件层面可以从两个方面来保证数据的安全性。 RAID组 节点冗余 这里节点冗余包括MDS和OSS节点，2个MDS节点互为active/stanby模式，两个OSS节点均为active
使用shell脚本一键搭建spark集群
04-02
#!/bin/bash # 定义变量 SPARK_HOME=/opt/spark SPARK_MASTER_URL=spark://master:7077 SPARK_WORKER_CORES=2 SPARK_WORKER_MEMORY=4g SPARK_WORKER_INSTANCES=2 ...echo "Spark集群已搭建完成！
评论
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值