Support Policies for RHEL Resilient Storage - Clustered LVM with lvm2-cluster's clvmd

Contents

• Overview
  • Applicable Environments
  • Useful References and Guides
  • Introduction
• Policies

Overview

Applicable Environments

• Red Hat Enterprise Linux (RHEL) 6 or 7 with the Resilient Storage Add-On

Useful References and Guides

• Support Policies for RHEL Resilient Storage Clusters

Introduction

This guide describes Red Hat's policies applicable to use of lvm2-cluster to manage clustered LVM volumes with clvmd. Users of clvmd should adhere to these policies in order to be eligible for support from Red Hat with the appropriate product support subscriptions.

Policies

Where these policies apply: There are multiple methods of managing LVM volumes in a High Availability or Resilient Storage cluster. Use of lvm2-cluster's clvmd functions is one such method. clvmd will be responsible for managing a particular volume group only if that volume group has the "clustered" attribute enabled in its metadata. This can be viewed with the vgs command, looking for the c character in the Attr column.

When these policies talk about applying to "clustered volume groups", it is referring specifically to LVM volume groups that have this clustered attribute enabled. If no such volume group exists with that attribute set, then clvmd effectively is not managing any device or volume, and has little impact on the operation of LVM - therefor these policies only apply if clvmd has volumes to manage.

If using a different method to manage LVM in a RHEL High Availability cluster, then Red Hat's general policies on LVM in a cluster may still apply:

• See: Support policies for RHEL High Availability clusters - LVM in a cluster

All nodes must share access to all clustered physical volumes and run clvmd: Any LVM physical volume (PV) that belongs to a clustered volume group must be accessible to every node of the cluster. The expected behavior around asymmetric storage access is not defined in clvmd's implementation, and Red Hat cannot provide assistance with such setups. If clvmd is to be used to manage clustered volume groups, the clvmd daemon should be configured to run on all member nodes of the cluster, and all nodes should have equal access to the clustered volume group PVs.

Conditions with RHEL 6, clvmd, and exclusive LVM resources managed by pacemaker: See the RHEL High Availability policy guide for details on conditions related to use of the LVM agent with exclusive clustered volume groups:

• See: Support policies for RHEL High Availability clusters - LVM in a cluster

clvmd replaced by lvmlockd in RHEL 8: lvm2-cluster (clvmd) is not provided in RHEL 8 Resilient Storage, as clustered LVM management is offered by lvmlockd in RHEL 8. See Support Policies for RHEL High Availability - LVM management in a cluster for further policy information.