Testing - Rocky Linux 9.5 Pacemaker Setup

I set up two Rocky Linux 9.5 VMs:
```
rocky-linux-95-1
rocky-linux-95-2
```
rocky-linux-95-1 will be my ISCSI target. I attached a 20GB disk to this VM.
rocky-linux-95-2 is my cluster nodes.
Ran the following steps on all nodes:
Switched to the root user:
```
sudo su -
```

Set FQDN hostnames on all nodes:

hostnamectl set-hostname rocky-linux-95-1.lvm.test
hostnamectl set-hostname rocky-linux-95-2.lvm.test

Updated the /etc/hosts file to make all nodes aware of each other:

cat << "EOF" | sudo tee /etc/hosts
192.168.1.48 rocky-linux-95-1.lvm.test rocky-linux-95-1
192.168.1.49 rocky-linux-95-2.lvm.test rocky-linux-95-2
EOF

Enabled the High Availability repository:

dnf config-manager --set-enabled highavailability

Installed these packages:

dnf install -y pacemaker
dnf install -y pcs

Started the pcsd service:
```
systemctl enable --now pcsd
```
Confirmed the pcsd service was running on all nodes:
```
systemctl status pcsd
```

Allowed high-availability services through firewalld:

firewall-cmd --add-service=high-availability --permanent 
firewall-cmd --reload 

Set the cluster admin password on all nodes:
```
passwd hacluster 
```
Ran the following steps on the rocky-linux-95-2 node:

Configured the basic cluster settings:

pcs host auth rocky-linux-95-1.lvm.test rocky-linux-95-2.lvm.test -u hacluster

Configured the cluster:

pcs cluster setup ha_cluster rocky-linux-95-1.lvm.test rocky-linux-95-2.lvm.test

Started cluster services:
```
pcs cluster start --all
```
Configured auto-start:
```
pcs cluster enable --all
```
The status looked good from the corosync side: ``` pcs status corosync

Membership information

Nodeid      Votes Name
     1          1 rocky-linux-95-1.lvm.test
     2          1 rocky-linux-95-2.lvm.test (local) ```

Testing - iSCSI Setup

All steps are ran on the rocky-linux-95-1 node
Switched to the root account:
```
sudo su -
```
Install the targetcli program:
```
dnf install -y targetcli 
```
Started and enabled the target service:
```
systemctl enable --now target
```
Ran targetcli:
```
targetcli
```
Changed directory into the fileio directory:
```
cd backstores/ramdisk
```
Created a 10GB backstore file:
```
create name=storage size=10G
```
Changed into the /iscsi directory:
```
cd /iscsi
```

Created the target:

create iqn.2025-03.test.lvm:rocky-linux-95-1

Changed into the iqn.2025-03.test.lvm:rocky-linux-95-1/tpg1/portals directory:
```
cd iqn.2025-03.test.lvm:rocky-linux-95-1/tpg1/portals
```
Created a portal to listen for incoming connections (if a message regarding the NetworkPortal already existing appears, this has already then been configured):
```
create 0.0.0.0 3260
```
Switched to the acls directory:
```
cd ../acls
```

Added rocky-linux-95-2 and rocky-linux-95-3 as initiators:

create iqn.2025-03.test.lvm.initiator:rocky-linux-95-2

Switched to the luns directory:
```
cd ../luns
```
Mapped the backstore to the iscsi target:
```
create /backstores/ramdisk/storage
```
Changed back to the root directory:
```
cd ../../../..
```
Saved the configuration:
```
saveconfig
```
Then exited the session:
```
exit
```

Allowed the iscsi-target traffic through firewalld:

firewall-cmd --add-service=iscsi-target --permanent
firewall-cmd --reload

These steps were ran on rocky-linux-95-2:
Installed the initiator utilities:
```
dnf install -y iscsi-initiator-utils 
```

Created the initiatorname.iscsi file:

cat << "EOF" | sudo tee /etc/iscsi/initiatorname.iscsi
InitiatorName=iqn.2025-03.test.lvm.initiator:rocky-linux-95-2
EOF

Discovered the node:

iscsiadm -m discovery -t sendtargets -p 192.168.1.48

Logged into the node successfully:

iscsiadm -m node --targetname iqn.2025-03.test.lvm:rocky-linux-95-1 --portal 192.168.1.48:3260 --login

Configured to log into the target on boot:

iscsiadm -m node --targetname iqn.2025-03.test.lvm:rocky-linux-95-1 --portal 192.168.1.48:3260 --op update --name node.startup --value automatic

Restarted the iscsid service:
```
sudo systemctl restart iscsid
```

Verified the iSCSI disk on rocky-linux-95-1 was found:

lsblk
NAME        MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS
sda           8:0    0    10G  0 disk

Testing - LVM Setup

Steps performed on all nodes:

Changed the LVM System ID to uname:

sed -i 's/# system_id_source = "none"/system_id_source = "uname"/' /etc/lvm/lvm.conf 

Steps performed on rocky-linux-95-2:

Set the label as gpt:

parted --script /dev/sda "mklabel gpt"

Assigned 100% of the space to the primary partition:

parted --script /dev/sda "mkpart primary 0% 100%"

Enabled LVM:

parted --script /dev/sda "set 1 lvm on"

Created a physical volume:
```
pvcreate /dev/sda1 
```

Created a volume group:

vgcreate --setautoactivation n vg_ha /dev/sda1

Confirmed the hostname and volume group match:

vgs -o+systemid

VG    #PV #LV #SN Attr   VSize    VFree   System ID                
rl      1   3   0 wz--n- <199.00g      0                           
vg_ha   1   0   0 wz--n-  <10.00g <10.00g rocky-linux-95-2.lvm.test

Created the logical volume:
```
lvcreate -l 100%FREE -n lv_ha vg_ha 
```
Formatted the volume as ext4:
```
mkfs.ext4 /dev/vg_ha/lv_ha 
```

Set the shared storage as a cluster resource:

pcs resource create lvm_ha ocf:heartbeat:LVM-activate vgname=vg_ha vg_access_mode=system_id group ha_group --future 

Confirmed the status:
```
pcs status
```

The shared LVM was then available:

lsblk
NAME            MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS
sda               8:0    0    10G  0 disk 
└─sda1            8:1    0    10G  0 part 
└─vg_ha-lv_ha 253:3    0    10G  0 lvm