Veritas Volume Manager and Veritas Cluster: RHCA Administration Guide

Veritas Volume Manager (VxVM) and Veritas Cluster Server (VCS) are enterprise storage and HA solutions used in large-scale data centers. RHCA candidates should understand Veritas concepts even if they primarily use LVM and Pacemaker.

Veritas vs Linux Native Tools

Veritas Volume Manager (VxVM)

VxVM Hierarchy

Physical Disks → Disk Groups → Volumes → Filesystems

# Physical disk (plex) — equivalent to LVM PV
# Disk Group (DG)      — equivalent to LVM VG
# Volume               — equivalent to LVM LV

VxVM Key Commands

# Initialize disk for VxVM:
# vxdisksetup -i /dev/sdb            # add disk to VxVM control

# Create disk group:
# vxdg init mydg /dev/sdb /dev/sdc

# Add disk to disk group:
# vxdg -g mydg adddisk /dev/sdd

# Create volume:
# vxassist -g mydg make myvol 10g

# Create mirrored volume (RAID-1):
# vxassist -g mydg make myvol 10g layout=mirror

# Create striped volume (RAID-0):
# vxassist -g mydg make myvol 10g layout=stripe

# List disk groups:
# vxdg list

# List volumes:
# vxprint -g mydg

# Extend volume:
# vxassist -g mydg growby myvol 5g

# Remove volume:
# vxedit -rf rm /dev/vx/rdsk/mydg/myvol

VxFS Commands

# Create VxFS filesystem:
# mkfs -F vxfs /dev/vx/rdsk/mydg/myvol

# Mount VxFS:
# mount -F vxfs /dev/vx/dsk/mydg/myvol /mnt/data

# Grow VxFS (online):
# fsadm -F vxfs -b 20g /mnt/data     # extend to 20G

# Snapshot:
# vxassist -g mydg snapshot myvol snapvol

Veritas Cluster Server (VCS)

VCS Concepts

• Service Group — group of resources that fail over together (like a Pacemaker resource group)
• Resource — individual component (IP, disk, application)
• Agent — monitors and manages a resource type
• HAD — High Availability Daemon — core VCS process

VCS Commands

# Check cluster status:
# hastatus -sum                       # summary
# hastatus -group                     # per service group

# List service groups:
# hagrp -list

# Online/offline service group:
# hagrp -online websg -sys node1
# hagrp -offline websg -sys node1

# Switch service group to other node:
# hagrp -switch websg -to node2

# Freeze (prevent failover):
# hagrp -freeze websg -persistent

# Unfreeze:
# hagrp -unfreeze websg -persistent

# List resources:
# hares -list

# Check resource state:
# hares -display
# hares -state

# Bring resource online:
# hares -online web_ip -sys node1

VCS Config Files

/etc/VRTSvcs/conf/config/main.cf    # main cluster config
/etc/VRTSvcs/conf/config/types.cf   # resource type definitions

# Validate config:
# hacf -verify /etc/VRTSvcs/conf/config/

Troubleshooting VCS

# Check VCS logs:
# cat /var/VRTSvcs/log/engine_A.log

# Check resource faults:
# hares -display -attribute Faulted

# Clear fault:
# hares -clear web_app -sys node1