Server Redundancy with Server Affinity Print

  • server, redundanz, affinity, server group, verfügbarkeit, openstack
  • 1

If you want to operate a server cluster or a redundant server within the ScaleUpOpen Cloud, it is important to make sure that your provisioned instances (VMs) are not running on the same host (physical hardware). Otherwise, your redundant setup would be offline in case of a hardware failure of a single physical host/server. During normal operations, the OpenStack platform decied where to place your provisioned instances (VMs). However, by using so called server groups you can influence where your instances are placed. Please follow these steps:

  1. With the help of the OpenStack CLI you need to create a server group. In this example we want to use the Policy Anti-Affinity in order to make sure that VMs are NOT placed on the same hardware. In order to do this, please use this command:
    openstack server group create Group01 --policy anti-affinity

  2. You should receive the following output:

    +----------+--------------------------------------+
    | Field | Value                                   |
    +----------+--------------------------------------+
    | id | a0bb8ea7-3348-42c7-a3da-e9593a8d7821       |
    | members |                                       |
    | name | Group01                                 |
    | policies | anti-affinity                        |
    +----------+--------------------------------------+



You have created a new server group with the name Group01. In the OpenStack Dashboard you can now use this server group when provisioning new instances.

Server Group Anti-Affinity

If you create additionel instances (VMs) with this server group, OpenStack wil make sure that instances in this grip will not be on the same physical hardware.

You can also create server groups with a Policy Affinity (--policy affinity), in order to group instances (VMs) on the same hardware.

Was this answer helpful?

« Back