gluster集群服务器IP地址更改后导致服务无法启动的一个解决方法

问题描述

原先建立了一个3主机的gluster集群，每个机器上包含2个bricks，总共6个bricks，采用4data加2EC的数据存放架构。某天实验室路由配置改变了，原先的三个主机IP全部发生变化，gluster启动后找不到peer，服务无法启动。

思路

属于一种hacker方法，将原有的配置文件里面所有旧的IP地址改为新的IP地址。目标文件都在目录/var/lib/glusterd/下面。

首先需要将目录里面所有文件包含IP字段的文本进行替换。

命令如下

for f in $(find ./ -type f); do echo $f && sed -i 's/192.168.111.51/192.168.120.20/g' $f ;done

其次要找到所有包含IP字段名字的文件，并且将文件名中的IP地段做同样的重命名

mv ./vols/gvfoo/bricks/192.168.111.50:-home-xqfs-gluster-b0            ./vols/gvfoo/bricks/192.168.120.10:-home-xqfs-gluster-b0
mv ./vols/gvfoo/bricks/192.168.111.50:-home-xqfs-gluster-b1            ./vols/gvfoo/bricks/192.168.120.10:-home-xqfs-gluster-b1
mv ./vols/gvfoo/bricks/192.168.111.50:-home-xqfs-gluster-b2            ./vols/gvfoo/bricks/192.168.120.10:-home-xqfs-gluster-b2
mv ./vols/gvfoo/bricks/192.168.111.50:-home-xqfs-gluster-b3            ./vols/gvfoo/bricks/192.168.120.10:-home-xqfs-gluster-b3
mv ./vols/gvfoo/bricks/192.168.111.51:-home-xqfs-gluster-b0            ./vols/gvfoo/bricks/192.168.120.20:-home-xqfs-gluster-b0
mv ./vols/gvfoo/bricks/192.168.111.52:-home-xqfs-gluster-b0            ./vols/gvfoo/bricks/192.168.120.21:-home-xqfs-gluster-b0
mv ./vols/gvfoo/gvfoo.192.168.111.50.home-xqfs-gluster-b0.vol               ./vols/gvfoo/gvfoo.192.168.120.10.home-xqfs-gluster-b0.vol
mv ./vols/gvfoo/gvfoo.192.168.111.50.home-xqfs-gluster-b1.vol               ./vols/gvfoo/gvfoo.192.168.120.10.home-xqfs-gluster-b1.vol
mv ./vols/gvfoo/gvfoo.192.168.111.50.home-xqfs-gluster-b2.vol               ./vols/gvfoo/gvfoo.192.168.120.10.home-xqfs-gluster-b2.vol
mv ./vols/gvfoo/gvfoo.192.168.111.50.home-xqfs-gluster-b3.vol               ./vols/gvfoo/gvfoo.192.168.120.10.home-xqfs-gluster-b3.vol
mv ./vols/gvfoo/gvfoo.192.168.111.51.home-xqfs-gluster-b0.vol               ./vols/gvfoo/gvfoo.192.168.120.20.home-xqfs-gluster-b0.vol
mv ./vols/gvfoo/gvfoo.192.168.111.52.home-xqfs-gluster-b0.vol               ./vols/gvfoo/gvfoo.192.168.120.21.home-xqfs-gluster-b0.vol
mv ./vols/gvfoo/gvfoo.client.192.168.111.50.home-xqfs-gluster-b0.vol ./vols/gvfoo/gvfoo.client.192.168.120.10.home-xqfs-gluster-b0.vol
mv ./vols/gvfoo/gvfoo.client.192.168.111.50.home-xqfs-gluster-b1.vol ./vols/gvfoo/gvfoo.client.192.168.120.10.home-xqfs-gluster-b1.vol
mv ./vols/gvfoo/gvfoo.client.192.168.111.50.home-xqfs-gluster-b2.vol ./vols/gvfoo/gvfoo.client.192.168.120.10.home-xqfs-gluster-b2.vol
mv ./vols/gvfoo/gvfoo.client.192.168.111.50.home-xqfs-gluster-b3.vol ./vols/gvfoo/gvfoo.client.192.168.120.10.home-xqfs-gluster-b3.vol
mv ./vols/gvfoo/gvfoo.client.192.168.111.51.home-xqfs-gluster-b0.vol ./vols/gvfoo/gvfoo.client.192.168.120.20.home-xqfs-gluster-b0.vol
mv ./vols/gvfoo/gvfoo.client.192.168.111.52.home-xqfs-gluster-b0.vol ./vols/gvfoo/gvfoo.client.192.168.120.21.home-xqfs-gluster-b0.vol

对集群中的三台机器做同样的操作，IP字段要做相应的更高。完成后，重启service。

systemctl start glusterd
systemctl start glusterfsd

最后，再把gluster里面对应的volume启动就可以了。