Gluster 安装

Gluster 安装

（镜像）

 

1.   环境准备

机器名	eth0	说明
server01	192.168.100.30/24	Centos 6.3 64bit ，多准备一块磁盘
server02	192.168.100.31/24	Centos 6.3 64bit，多准备一块磁盘
Client	192.168.100.60/24	Centos 6.3 64bit

 

2.   格式磁盘并挂载（在两个节点上均做如下设置）

安装mkfs.xfs命令包

yum install xfsprogs  

mkfs.xfs -i size=512 /dev/sdb 

mkdir -p /export/brick1

修改分区表 vi /etc/fstab 末尾添加

/dev/sdb        /export/brick1                  xfs     defaults        1 2

mount –a &&mount

 

3.   安装gluster（所有节点均需安装）

添加源

wget -P /etc/yum.repos.d http://download.gluster.org/pub/gluster/glusterfs/LATEST/EPEL.repo/glusterfs-epel.repo

安装gluster

yum install glusterfs{-fuse,-server}

启动服务

service glusterd start

service gluster status

 

4.   配置gluster

注：以下配置在一个节点上操作即可

配置对端信任节点（即镜像服务器）

gluster peer probe 192.168.100.31

创建一个功能为镜像的集群卷（可以使用域名）

gluster volume create gv0 replica 2 192.168.100.30:/export/brick1 192.168.100.31:/export/brick1
gluster volume start gv0

查看卷信息

 [root@server01 brick1]# gluster volume info

Volume Name: gv0

Type: Replicate

Volume ID: 3bf26a6a-63ce-44ed-bd57-fffd52352130

Status: Started

Number of Bricks: 1 x 2 = 2

Transport-type: tcp

Bricks:

Brick1: 192.168.100.30:/export/brick1

Brick2: 192.168.100.31:/export/brick1

 

5.   验证

在客户端192.168.100.60上验证gluster，

1.      安装gluster也就是步骤3的server组件就包含了客户端

2.      只安装客户端组件

yum -y install fuse fuse-libs

yum install glusterfs glusterfs-fuse glusterfs-rdma

自动挂载
修改/etc/fstab 文件
server1:/test-volume /mnt/glusterfs glusterfs defaults,_netdev 0 0
192.168.100.30:/gv0 /mnt glusterfs defaults,netdev 0 0 

mount -t glusterfs 192.168.100.30:/gv0 /mnt

[root@node mnt]#  mount -l 

/dev/mapper/VolGroup-lv_root on / type ext4 (rw)

proc on /proc type proc (rw)

sysfs on /sys type sysfs (rw)

devpts on /dev/pts type devpts (rw,gid=5,mode=620)

tmpfs on /dev/shm type tmpfs (rw,rootcontext="system_u:object_r:tmpfs_t:s0")

/dev/sda1 on /boot type ext4 (rw)

none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw)

sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw)

/dev/sr0 on /media/CentOS_6.3_Final type iso9660 (ro,nosuid,nodev,uhelper=udisks,uid=500,gid=500,iocharset=utf8,mode=0400,dmode=0500) [CentOS_6.3_Final]

/dev/sr0 on /media/cdrom type iso9660 (ro) [CentOS_6.3_Final]

192.168.100.30:/gv0 on /mnt type fuse.glusterfs (rw,default_permissions,allow_other,max_read=131072)

 

已经将gv0挂载至/mnt目录，在客户端目录下创建100个文件

for i in `seq -w 1 100`; do cp -rp /var/log/messages  /mnt/copy-test-$i; done

[root@node mnt]# ls

copy-test-001  copy-test-013

[root@node mnt]# ls -lA /mnt | wc -l

101

到服务器192.168.100.30上查看验证

[root@server01 brick1]# ls

copy-test-001  copy-test-013

[root@server01 brick1]# ls -l  | wc -l

101

服务器192.168.100.31上查看验证

[root@server02 brick1]# ls

copy-test-001  copy-test-013  

[root@server02 brick1]#  ls -l  | wc -l

101

说明：上述例子用2台服务器上的一块硬盘做了镜像冗余，客户端上传文件会自动同步到2台服务器的gluster卷上。

 

1.   故障恢复验证

查看当前挂载情况，挂载的是server01的地址

[root@node mnt]# mount 

192.168.100.30:/gv0 on /mnt type fuse.glusterfs (rw,default_permissions,allow_other,max_read=131072)

查看当前数据server01和server02里均为空

在客户端创建测试数据，数据同步到2台服务器上

[root@server02 brick1]# ls

copy-test-001  copy-test-013  copy-test-025

由于当前mount的地址为server01的地址，现在当server01断开网络后，验证客户端是否仍然可以与server02的卷交互保障业务不中断。

[root@node mnt]# ping 192.168.100.30

PING 192.168.100.30 (192.168.100.30) 56(84) bytes of data.

From 192.168.100.60 icmp_seq=1 Destination Host Unreachable

^C

--- 192.168.100.30 ping statistics ---

3 packets transmitted, 0 received, +1 errors, 100% packet loss, time 2423ms

[root@node mnt]# ls

copy-test-001  copy-test-013  copy-test-025

由此可见当server01断开网络后，客户端仍然可以与server02交互数据。

此时在客户端删除并更新数据，server02仍然是断开前的数据，恢复后是否可以从server01同步回数据。

在客户端删除创建的测试数据，并新建文件test，server01跟进同步了

[root@node mnt]# rm -rf * 

[root@node mnt]# touch test

[root@node mnt]# ls

Test

[root@node mnt]# netstat -antulp |grep 192.168.*

tcp 0  52 192.168.100.60:22  192.168.100.1:49994         ESTABLISHED 2106/sshd           

tcp 0 1 192.168.100.60:1023   192.168.100.30:24009        SYN_SENT    2358/glusterfs      

tcp 0 1 192.168.100.60:1022   192.168.100.30:24007        SYN_SENT    2358/glusterfs      

tcp 0 0 192.168.100.60:1021  192.168.100.31:24009        ESTABLISHED 2358/glusterfs   

[root@server02 brick1]# ls

Test

server01里由于网络故障，导致数据无法更新

[root@server01 brick1]# ls

copy-test-001  copy-test-013

网络恢复后，间隔一段时间，发现server02同步了server01的数据

[root@server01 brick1]# ls

Test

[root@node mnt]# netstat -antulp |grep 192.168.*

tcp        0     52 192.168.100.60:22           192.168.100.1:49994         ESTABLISHED 2106/sshd           

tcp        0      0 192.168.100.60:1023         192.168.100.30:24009        ESTABLISHED 2358/glusterfs      

tcp        0      0 192.168.100.60:1022         192.168.100.30:24007        ESTABLISHED 2358/glusterfs      

tcp        0      0 192.168.100.60:1021         192.168.100.31:24009        ESTABLISHED 2358/glusterfs  

 

由此可见当服务器中有一台故障后，客户端的业务仍然可以继续，不会受到影响，故障服务器恢复后，会从没有发生故障的服务器同步数据，来确保数据的一致性。

 

7 扩展gluster volume

Server01和server02均再增加一块10G大小的硬盘，原有的是20G的硬盘互相镜像，现在为gluster卷增加10G大小的空间

[root@server01 ~]# mkfs.xfs -i size=512 /dev/sdc

mkdir -p /export/brick2

编辑vi/etc/fstab 末尾添加

/dev/sdc                /export/brick2          xfs     defaults       1 2

原卷空间大小

[root@node /]# df -h /mnt

Filesystem            Size  Used Avail Use% Mounted on

192.168.100.30:/gv0    20G   33M   20G   1% /mnt

Gluster卷情况

[root@server01 ~]# gluster volume info

Volume Name: gv0

Type: Replicate

Volume ID: 3bf26a6a-63ce-44ed-bd57-fffd52352130

Status: Started

Number of Bricks: 1 x 2 = 2

Transport-type: tcp

Bricks:

Brick1: 192.168.100.30:/export/brick1

Brick2: 192.168.100.31:/export/brick1

增加新建的10G大小的卷

# gluster volume add-brick gv0 192.168.100.30:/export/brick2 192.168.100.31:/export/brick2

[root@server01 brick1]# gluster volume info

Volume Name: gv0

Type: Distributed-Replicate

Volume ID: 3bf26a6a-63ce-44ed-bd57-fffd52352130

Status: Started

Number of Bricks: 2 x 2 = 4

Transport-type: tcp

Bricks:

Brick1: 192.168.100.30:/export/brick1

Brick2: 192.168.100.31:/export/brick1

Brick3: 192.168.100.30:/export/brick2

Brick4: 192.168.100.31:/export/brick2

客户端查看扩容后的卷大小

[root@node /]# df -h /mnt

Filesystem            Size  Used Avail Use% Mounted on

192.168.100.30:/gv0    30G   65M   30G   1% /mnt

至此顺利的完成了对gluster镜像卷的扩容。

8.   移除卷

移除卷会将卷上的数据迁移到其他可用的卷上

gluster volume remove-brick gv0192.168.100.32:/export/brick1 192.168.100.33:/export/brick1 start

gluster volume remove-brick gv0192.168.100.32:/export/brick1 192.168.100.33:/export/brick1 commit

查看信息发现server3和server4的卷已经移除，并且卷上的数据未丢失,已迁移到可用的卷上。

gluster> volume info

Volume Name: gv0

Type: Distributed-Replicate

Volume ID:3bf26a6a-63ce-44ed-bd57-fffd52352130

Status: Started

Number of Bricks: 2 x 2 = 4

Transport-type: tcp

Bricks:

Brick1: 192.168.100.30:/export/brick1

Brick2: 192.168.100.31:/export/brick1

Brick3: 192.168.100.30:/export/brick2

Brick4: 192.168.100.31:/export/brick2

9.   访问限制

通过访问限制，允许某些特定主机访问存储，拒绝某些特定主机访问存储

gluster volume set gv0 auth.allow192.168.100.32,192.168.100.60

gluster volume set gv0 auth.reject <ip-address>

gluster> volume info

Volume Name: gv0

Type: Distributed-Replicate

Volume ID:3bf26a6a-63ce-44ed-bd57-fffd52352130

Status: Started

Number of Bricks: 2 x 2 = 4

Transport-type: tcp

Bricks:

Brick1: 192.168.100.30:/export/brick1

Brick2: 192.168.100.31:/export/brick1

Brick3: 192.168.100.30:/export/brick2

Brick4: 192.168.100.31:/export/brick2

Options Reconfigured:

auth.reject: none

auth.allow: 192.168.100.32,192.168.100.60

diagnostics.latency-measurement: on

diagnostics.count-fop-hits: on

 

10.             单机做gluster

Server03 有2块硬盘sdb、sdc希望通过gluster在这两块硬盘之间做镜像备份

格式化两块磁盘并挂载，如下：

/dev/sdc on /export/brick2 type xfs (rw)

/dev/sdb on /export/brick1 type xfs (rw)

 

gluster volume create gv0 replica 2192.168.100.32:/export/brick1 192.168.100.32:/export/brick2

Multiple bricks of a replicate volume arepresent on the same server. This setup is not optimal.

Do you still want to continue creating thevolume?  (y/n) y

Creation of volume gv0 has been successful.Please start the volume to access data.

[root@server03 /]# gluster volume start gv0

Starting volume gv0 has been successful

[root@server03 /]# gluster volume status

Status of volume: gv0

Gluster process                                        Port    Online  Pid

------------------------------------------------------------------------------

Brick 192.168.100.32:/export/brick1                     24009   Y      2252

Brick 192.168.100.32:/export/brick2                     24010   Y      2259

NFS Server on localhost                                 38467   Y      2265

Self-heal Daemon on localhost                           N/A     Y      2271

 

去客户端挂载测试

[root@node /]# mount -t glusterfs192.168.100.32:/gv0 /volume/

[root@node /]# mount

192.168.100.32:/gv0 on /volume typefuse.glusterfs (rw,default_permissions,allow_other,max_read=131072)