Setup Proxmox Cluster High Availability

from the main node issue the command

pvecm create clusterone 

from the other node :

pvecm add MAIN_NODE_IP 

we got cluster running now .

High Availability :

click on the 'Datacenter' --> HA --> Add : choose the virtual machine you need to add to HA.

to test HA is running  or not , go to the node where the virtual machine is install and down the network interface :

ifdown eth0 

the virtual machine should be still running while the main node is off.

Setting up Gluster
Reference 1
Reference 2
Refernce 3
Notice : Directory Need to be Manipulated to serve this job check Reference 2

 Add the GPG key to apt:



    wget -O - http://download.gluster.org/pub/gluster/glusterfs/3.11/rsa.pub | apt-key add -



Add the source:



    echo deb http://download.gluster.org/pub/gluster/glusterfs/3.11/LATEST/Debian/jessie/apt jessie main > /etc/apt/sources.list.d/gluster.list 



or:

    echo deb http://download.gluster.org/pub/gluster/glusterfs/3.11/LATEST/Debian/stretch/apt stretch main > /etc/apt/sources.list.d/gluster.list 



Update package list:



    apt-get update



Install:



    apt-get install  glusterfs-server glusterfs-client 

from the master node

  gluster peer probe node1.proxmox
 

create the share named testvol with two replicas (please note that the number of replicas is equal to the number of servers :

 gluster volume create testvol replica 2 transport tcp master.proxmox:/data/testvol node1.proxmox:/data/testvol force

Gluster may complainn of spli brain issue , but I think proxmox can avoid that using fencing or Watchdog. Start the Volume

gluster volume start testvol

you can check the connection is established or not using :

 netstat -tap | grep glusterfsd 

if not check the required ports are open .
Now , from Proxmox Datacenter node , Choose --> Storage --> Add --> GlusterFS enter the name you like for this gluster storage, the ip of the two servers and the volume name "testvol" in our case

Don't forget to enables watchdog

just uncomment the line

# select watchdog module (default is softdog)
WATCHDOG_MODULE=ipmi_watchdog

Now It is time to do security job to secure the system

Commands:
Start Stop the Gluster service

 service glusterfs-server start | stop | status 

Information about gluster Volume

gluster volume info

Check status of high availability

systemctl status pve-ha-crm.service 

Check the Cluster Status

 pvecm status 

High Availability Command:

USAGE: ha-manager  [ARGS] [OPTIONS]
       ha-manager groupadd  -nodes  [OPTIONS]
       ha-manager groupconfig
       ha-manager groupremove 
       ha-manager groupset  [OPTIONS]

       ha-manager add  [OPTIONS]
       ha-manager config  [OPTIONS]
       ha-manager migrate  
       ha-manager relocate  
       ha-manager remove 
       ha-manager set  [OPTIONS]

       ha-manager status  [OPTIONS]

       ha-manager help [] [OPTIONS]