To be perfected...

Experimental environment

Host (IP)	service
server1(172.25.11.1)	mfsmaster corosync+pacemaker
server2(172.25.11.2)	chunk server
server3(172.25.11.3)	chunk server
server4(172.25.11.4)	mfsmaster corosync+pacemaker
Real machine (172.25.11.250)	client

pacemaker+corosync for high availability

Deployment of server1 and server4 host mfs master

• Let's take the installation of server1 as an example. The installation of server4 is the same.
• Download and install.

[root@server1 ~]# ls
moosefs-cgi-3.0.103-1.rhsystemd.x86_64.rpm
moosefs-cgiserv-3.0.103-1.rhsystemd.x86_64.rpm
moosefs-chunkserver-3.0.103-1.rhsystemd.x86_64.rpm
moosefs-cli-3.0.103-1.rhsystemd.x86_64.rpm
moosefs-client-3.0.103-1.rhsystemd.x86_64.rpm
moosefs-master-3.0.103-1.rhsystemd.x86_64.rpm
moosefs-metalogger-3.0.103-1.rhsystemd.x86_64.rpm

[root@server1 ~]# yum install moosefs-master-3.0.103-1.rhsystemd.x86_64.rpm   moosefs-cli-3.0.103-1.rhsystemd.x86_64.rpm  moosefs-cgiserv-3.0.103-1.rhsystemd.x86_64.rpm   moosefs-cgi-3.0.103-1.rhsystemd.x86_64.rpm -y

• Serr1 add parsing

[root@server1 ~]# vim /etc/hosts
172.25.11.1     server1  mfsmaster

• Open the service and view the port

[root@server1 mfs]# systemctl start moosefs-master
[root@server1 mfs]# netstat -antlp
[root@server1 mfs]# systemctl start moosefs-cgiserv
[root@server1 mfs]# netstat -antlp

• Access View in Browser: http://172.25.11.1:9425/mfs.cgi
  

Serr1 and server4 configure highly available yum sources

• Taking server1 as an example, the construction of server4 is the same

[root@server1 system]# vim /etc/yum.repos.d/rhel7.repo

[rhel7.3]
name=rhel7.3
gpgcheck=0
baseurl=http://172.25.11.250/rhel7.3

[HighAvailability]
name=HighAvailability
gpgcheck=0
baseurl=http://172.25.11.250/rhel7.3/addons/HighAvailability

[ResilientStorage]
name=ResilientStorage
gpgcheck=0
baseurl=http://172.25.11.250/rhel7.3/addons/ResilientStorage

Install pacemaker+corosync on server1 and server4

• server1:

[root@server1 system]# yum install pacemaker corosync -y

• server4:

[root@server4 ~]# yum install pacemaker corosync -y

Because of high availability, there is a need for privacy-free between server1 and server4

[root@server1 system]# ssh-keygen
[root@server1 system]# ssh-copy-id server4
[root@server1 system]# ssh server4
[root@server1 system]# ssh-copy-id server1
[root@server1 ~]# ssh server1

Install resource management tools on server1 and server4 and open corresponding services

• Serr1 is the same as server4. Here we take server1 as an example.

[root@server1 ~]# yum install pcs -y
[root@server1 ~]# systemctl start pcsd
[root@server1 ~]# systemctl enable pcsd
[root@server1 ~]# id hacluster
uid=189(hacluster) gid=189(haclient) groups=189(haclient)
[root@server1 ~]# passwd hacluster

Create mfs cluster on server 1 and start

• Creating Clusters

[root@server1 ~]# pcs cluster auth server1 server4
Username: hacluster
Password: 
server4: Authorized
server1: Authorized

• Name the cluster

[root@server1 ~]# pcs cluster setup --name mycluster server1 server4

• Open Cluster

[root@server1 ~]# pcs cluster start --all
server1: Starting Cluster...
server4: Starting Cluster...

View cluster status

[root@server1 ~]# corosync-cfgtool -s
[root@server1 ~]# pcs status corosync

Check and resolve errors

[root@server1 ~]# crm_verify -L -V
[root@server1 ~]# pcs property set stonith-enabled=false
[root@server1 ~]# crm_verify -L -V

Creating Cluster Resources and Adding VIP

• On server 1:

[root@server1 ~]# pcs resource create --help
[root@server1 ~]# pcs resource create vip ocf:heartbeat:IPaddr2 ip=172.25.11.100 cidr_netmask=32  op monitor interval=30s
[root@server1 ~]# pcs resource show
[root@server1 ~]# pcs status

• View vip on server 4:

[root@server4 ~]# crm_mon

test

• View vip on server 1 at this time:
  
• Turn off server 1 in cluster cluster cluster, and vip drifts to server 4.

[root@server1 ~]# pcs cluster stop server1

• Viewing the monitoring information shows that only server4 is online at this time:

[root@server4 ~]# crm_mon

• Open server1 in the cluster again and find vip is still on server4, but server1 and server4 are online when monitoring information is viewed.

[root@server1 ~]# pcs cluster start server1
server1: Starting Cluster...

[root@server4 ~]# crm_mon 
[root@server4 ~]# ip a

Overall, high availability deployment is successful!