新增三台centos7机器，最小化安装
ip分配是 服务端192.168.1.7 192.168.1.8 客户端192.168.1.9
主机名分别为 gluster01 gluster02 client01
hostnamectl set-hostname gluster01
hostnamectl set-hostname gluster02
hostnamectl set-hostname client01
配置所有机器网络，要求能访问外网（略）
关闭防火墙和selinux
systemctl stop firewalld
systemctl disable firewalld
setenforce 0
sed -i 's/enforcing/disabled/g' /etc/selinux/config
修改所有机器的hosts文件，添加对应的ip主机名解析
vi /etc/hosts
192.168.1.7 gluster01
192.168.1.8 gluster02
192.168.1.9 client01
配置yum源
cd /etc/yum.repos.d/
wget http://mirrors.aliyun.com/repo/Centos-7.repo
wget http://mirrors.aliyun.com/repo/epel-7.repo
yum -y install epel-release
安装服务端
yum install centos-release-gluster -y
yum install -y glusterfs glusterfs-server glusterfs-fuse
systemctl start glusterd
systemctl enable glusterd
在节点gluster01上，配置整个GlusterFS集群，把各个节点加入到集群
gluster peer probe gluster01
gluster peer probe gluster02
查看集群状态
gluster peer status
在两个服务节点上创建数据存储目录
mkdir -p /usr/local/share/models
在gluster01上创建GlusterFS磁盘
加上replica 2 就是2个节点中，每个节点都要把数据存储一次，就是一个数据存储2份，每个节点一份
如果不加replica 2，就是2个节点的磁盘空间整合成一个硬盘
gluster volume create models replica 2 gluster01:/usr/local/share/models gluster02:/usr/local/share/models force
启动集群
gluster volume start models
安装客户端
yum install -y glusterfs glusterfs-fuse
mkdir -p /mnt/models
挂载
mount -t glusterfs -o rw gluster01:models /mnt/models/
df查看
文件系统 1K-块 已用 可用 已用% 挂载点
/dev/sda2 18244432 1012448 16282176 6% /
devtmpfs 491416 0 491416 0% /dev
tmpfs 500680 0 500680 0% /dev/shm
tmpfs 500680 6792 493888 2% /run
tmpfs 500680 0 500680 0% /sys/fs/cgroup
/dev/sda1 194235 95079 84820 53% /boot
tmpfs 100136 0 100136 0% /run/user/0
gluster01:models 18244352 1012480 16282112 6% /mnt/model
其他操作命令
删除GlusterFS磁盘
# gluster volume stop models 先停止
# gluster volume delete models 再删除
卸载GlusterFS磁盘
gluster peer detach gluster02
ACL访问控制
gluster volume set models auth.allow 192.168.1.*,192.168.2.*
添加GlusterFS节点
# gluster peer probe gluster03
# gluster peer probe gluster04
# gluster volume add-brick models gluster03:/data/gluster gluster04:/data/gluster
迁移GlusterFS数据
# gluster volume remove-brick models gluster01:/usr/local/share/models gluster03:/usr/local/share/models start
# gluster volume remove-brick models gluster01:/usr/local/share/models gluster03:/usr/local/share/models status
# gluster volume remove-brick models gluster01:/usr/local/share/models gluster03:/usr/local/share/models commit
修复GlusterFS数据(在节点1宕机的情况下)
# gluster volume replace-brick models gluster01:/usr/local/share/models gluster03:/usr/local/share/models commit -force
# gluster volume heal models full
本文转自super李导51CTO博客，原文链接：http://blog.51cto.com/superleedo/2056732 ，如需转载请自行联系原作者
