Support Policies for RHEL High Availability Clusters - OpenStack Virtual Machines as Cluster Members

Updated -

Contents

Overview

Applicable Environments

  • Red Hat Enterprise Linux (RHEL) with the High Availability Add-On
  • Red Hat OpenStack to provide Virtual Machines that will be used as Cluster Members.

Useful References and Guides

Introduction

This guide offers Red Hat's policies, requirements, and limitations applicable to the use of OpenStack virtual machines as members of a RHEL High Availability cluster. Users of RHEL High Availability software components should adhere to these policies by installing only on the approved platforms in order to be eligible to receive assistance from Red Hat Support with the appropriate support subscriptions.

Policies

Consider general conditions for support of RHEL High Availability in virtualization environments

Support for OpenStack VMs as RHEL High Availability members or as Pacemaker Remote Nodes:

  • RHEL 8.7:
    • RHBA-2022:7573 for pacemaker-2.1.4-5.el8 or later.
    • RHBA-2022:7443 for resource-agents-4.9.0-29.el8 or later.
    • RHBA-2022:7520 for fence-agents-openstack-4.2.1-103.el8 or later with includes /usr/sbin/fence_openstack.
  • RHEL 9:
    • RHBA-2022:7937 with the following package(s): pacemaker-2.1.4-5.el9 or later.
    • RHBA-2022:7932 with the following package(s): resource-agents-paf-4.10.0-23.el9, resource-agents-4.10.0-23.el9, resource-agents-cloud-4.10.0-23.el9 or later.
    • The fence-agent is included in all versions of RHEL 9.

Limitations for OpenStack VMs as RHEL High Availability members or as Pacemaker Remote Nodes:

  • Red Hat only supports up to 8 nodes in a cluster of RHOSP Virtual Machines.
  • Red Hat only supports running RHEL-HA cluster on top of Red Hat OpenStack (version 16 or later).
  • Openstack-cinder-volume does not currently support “multiattach” mode. Multiattach mode means that a virtual disk can be attached to multiple cluster nodes at the same time. Currently the virtual disk can only be attached to one cluster node at a time.

Other considerations for OpenStack VMs as RHEL High Availability members or as Pacemaker Remote Nodes:

  • OpenStack has a default limit of 120 API calls per minute per IP (unless custom configurations have been used during OpenStack deployment). When configuring a very high amount of resources within the same cluster it is possible, even if unlikely, to hit the rate limit. When hitting the rate limit, some operations can timeout, and potentially lead to resource failures and fencing, as it's possible that some operations might just take a much longer time than expected. It is not possible to change the rate limit without OpenStack admin custom configuration. If you expect to run a large number of resources, then it's best to contact the OpenStack administrators ahead of time.

Comments