The guide walks through adding a new disk to a CentOS 8/RHEL 8 node that already runs VxVM, covering everything from RPM installation to mounting the filesystem and running basic health checks. It begins by installing the veritas‑vxvm and veritas‑vxfs packages with dnf, then loading the vxvm and vxfs kernel modules and confirming the driver via dmesg. Once the driver is verified, you create a physical volume on /dev/sdb, extend or build a volume group called vgdata, carve out a 50 GB logical volume named lvapp, format it with mkfs.vxfs, and mount it at /mnt/appdata while adding an fstab entry for persistence. Finally, the tutorial runs vxdisk to check disk health and performs a simple write‑read test to ensure everything is functioning correctly before declaring the setup complete.

Configure Veritas Storage on CentOS 8 or RHEL 8 Step‑by‑Step

If you’re adding a new disk to a production node that already runs the Veritas Volume Manager (VxVM) stack, this walk‑through will get your VxFS filesystem up and running in minutes. I’ve done this on a few servers after a bad driver update wiped out my volume group, so expect no fluff—just what you need.

1 – Install the Veritas Packages

Before you can do anything else you have to get the right RPMs on the box.

sudo dnf install veritas-vxvm veritas-vxfs

Why that matters: VxVM and VxFS are separate packages; if you skip one the whole stack falls apart when you try to create a volume.

2 – Load the Kernel Modules

The kernel must know about Veritas.

sudo modprobe vxvm
sudo modprobe vxfs

If those commands fail, check /etc/modprobe.d/ for an older vxvm.conf that blocks loading.

3 – Verify the Driver is Active

Run:

dmesg | grep -i vxvm

You should see a line like “vxvm: driver initialized”. If you only get warnings, there’s a version mismatch between your kernel and the Veritas package.

4 – Create a Physical Volume (PV)

Assume the new disk is /dev/sdb.

sudo pvcreate /dev/sdb

Why: VxVM works with PVs, not raw disks. The command writes the PV header and checks for bad sectors before you lose data.

5 – Build a Volume Group (VG)

If you already have a VG called vgdata, add the new PV to it; otherwise create one.

sudo vgcreate vgdata /dev/sdb          # only if creating
# or
sudo vgextend vgdata /dev/sdb          # adding

I’ve seen this fail when the disk already contains a legacy partition table; pvcreate -ff forces overwrite, but use it with caution.

6 – Create a Logical Volume (LV)

Choose how much space you want.

sudo lvcreate -L 50G -n lvapp vgdata

The -L flag limits the LV to 50 GB; use -l ALL if you want everything.

7 – Format the LV with VxFS

sudo mkfs.vxfs /dev/vgdata/lvapp -y

The -y skips confirmation. If you see “VxFS: no version 1” it means the disk is too new for the installed Veritas package—upgrade the RPM.

8 – Mount the Filesystem

Pick a mount point:

sudo mkdir /mnt/appdata
sudo mount -t vxfs /dev/vgdata/lvapp /mnt/appdata

Add it to /etc/fstab for persistence:

/dev/vgdata/lvapp /mnt/appdata vxfs defaults,nofail 0 0

9 – Check Disk Health

Use Veritas’s built‑in checker:

sudo vxdisk -d /dev/sdb -t

If it reports errors, you’ll need to replace the disk before moving on.

10 – Verify Everything Is Working

Write a test file and read it back:

echo "test" | sudo tee /mnt/appdata/test.txt
cat /mnt/appdata/test.txt

If that succeeds, you’re good to go. If not, look at dmesg for VxVM or VxFS errors.

That’s the bare minimum to get a new Veritas volume up and running on CentOS 8/RHEL 8. For more advanced features—snapshots, mirroring, or clustering—consult the official Veritas docs.