[1]
Packages
- Will be using two t3a.larg instances in an active/passive config.
- Package installation
sudo yum upgrade -y && sudo reboot -h now
sudo yum install pcs pacemaker fence-agents-all -y
Create Cluster
sudo passwd hacluster
Changing password for user hacluster.
New password:
Retype new password:
passwd: all authentication tokens updated successfully.
sudo systemctl enable --now pcsd.service
sudo pcs cluster auth ip-192-168-17-132.us-east-2.compute.internal ip-192-168-17-164.us-east-2.compute.internal --force
sudo pcs cluster setup --start --name nfs_cluster ip-192-168-17-132.us-east-2.compute.internal ip-192-168-17-164.us-east-2.compute.internal --force
- Here I disable stonith since I am performing an active/passive config. This is probably a very bad idea[3][4]
sudo pcs property set stonith-enabled=false
Volume Creation
- Create lvm, mount point, and deactivate the volume group[5].
sudo pvcreate /dev/nvme1n1
sudo vgcreate data /dev/nvme1n1
sudo lvcreate --size 195G --name nfs data
sudo mkfs.ext4 /dev/mapper/data-nfs
sudo mkdir -p /mnt/nfs
sudo vgchange -an data
0 logical volume(s) in volume group "data" now active
- Configure LVM for clustering[6]
- Add 'volume_list = []' to /etc/lvm/lvm.conf
sudo lvmconf --enable-halvm --services --startstopservices
sudo pcs resource create nfs LVM volgrpname=data exclusive=true --group nfs_cluster
sudo pcs resource create nfsshare Filesystem device=/dev/data/nfs directory=/mnt/nfs fstype=ext4 --group nfs_cluster
