Availability Zones

Overview

(AZs) are logical partitions of the Xloud Compute cluster that map to independent physical failure domains — separate racks, power feeds, or network segments. Distributing instances across multiple zones ensures workloads remain available when a localized hardware or network failure occurs.

Prerequisites

An active Xloud project with at least one running or available compute host
member or admin role assigned in your project
CLI users: Xloud CLI configured and credentials sourced

How Availability Zones Work

Each compute host is assigned to exactly one availability zone at deployment time. When you launch an instance and specify an AZ, the scheduler restricts placement to hosts within that zone. If no zone is specified, the scheduler selects the optimal host from any available zone.

Zone Attribute	Description
Zone Name	Unique identifier used when launching instances (e.g., `az1`, `nova`)
Zone Status	`available` — zone is healthy and accepting placements
Host Count	Number of compute hosts assigned to this zone (admin view only)
Default Zone	`nova` — the implicit zone when no preference is specified

The default zone name is nova. If your deployment defines custom zones (e.g., az-rack-1, az-rack-2), use those names when launching instances requiring fault isolation.

List Available Zones

Dashboard
CLI

Availability zone selection is surfaced in the Launch Instance dialog rather than a dedicated management page.

Open the Launch Instance dialog

Navigate to Project → Compute → Instances and click Launch Instance.

View available zones

On the Details tab, open the Availability Zone dropdown. All zones available to your project are listed with their current status.

Zones with status available accept new instance placements.

Authenticate

Source your credentials file to authenticate with the Xloud platform:

Load credentials

source admin-openrc.sh

Download the OpenRC file from Xloud Dashboard → Project → API Access → Download OpenStack RC File.

List all zones

List availability zones

openstack availability zone list --long

Expected output:

Sample output

+-----------+-------------+-----------+
| Zone Name | Zone Status | Hosts     |
+-----------+-------------+-----------+
| nova      | available   | -         |
| az-rack-1 | available   | -         |
| az-rack-2 | available   | -         |
+-----------+-------------+-----------+

The Hosts column is visible only to administrators. Regular project members see - regardless of the number of hosts in the zone.

Launch an Instance into a Specific Zone

Dashboard
CLI

Open the Launch Instance dialog

Navigate to Project → Compute → Instances and click Launch Instance.

Select the availability zone

On the Details tab, select the target zone from the Availability Zone dropdown.

Complete instance configuration

Continue through the remaining tabs (Source, Flavor, Networks, Security Groups, Key Pair) and click Launch Instance.

The instance is placed on a host within the selected zone. Verify in the instance details under Availability Zone.

Launch instance in a specific zone

openstack server create \
  --image <IMAGE_ID_OR_NAME> \
  --flavor <FLAVOR_NAME> \
  --network <NETWORK_ID_OR_NAME> \
  --key-name <KEYPAIR_NAME> \
  --availability-zone az-rack-1 \
  my-instance

Omit --availability-zone to let the scheduler choose the optimal host automatically.

Verify zone placement

openstack server show my-instance -c "OS-EXT-AZ:availability_zone"

The output shows the zone name confirming correct placement.

Relationship to Host Aggregates

Availability zones are a user-facing abstraction built on top of host aggregates. Administrators assign compute hosts to aggregates and optionally expose those aggregates as availability zones.

Concept	Audience	Purpose
Host Aggregate	Administrators	Group hosts by hardware capability or location
Availability Zone	End users	Select fault domain at instance launch
Host	Administrators	Individual compute node within an aggregate

A single host aggregate can be exposed as an availability zone by setting the availability_zone metadata key on the aggregate. Users see the zone name; the underlying aggregate membership is not exposed.

Best Practices

Zone distribution recommendations

Stateless tiers (web servers, API nodes): Distribute across at least two availability zones.
Stateful services (databases, message queues): Use server groups with anti-affinity policy within or across zones.
Single-instance workloads: Zone selection is optional — the scheduler ensures optimal placement.
Active-passive pairs: Place the primary in az-rack-1 and the standby in az-rack-2 to survive rack-level failures.

Next Steps

Launch an Instance

Step-by-step guide to launching instances with zone, flavor, and network configuration

Server Groups

Enforce affinity and anti-affinity placement policies across instances

Compute User Guide

Overview of all compute operations and instance lifecycle management

Compute Admin Guide

Configure host aggregates, zones, and cluster-level compute settings

Core Services

Other Services

Availability Zones

Overview

How Availability Zones Work

List Available Zones

Open the Launch Instance dialog

View available zones

Authenticate

List all zones

Launch an Instance into a Specific Zone

Open the Launch Instance dialog

Select the availability zone

Complete instance configuration

Relationship to Host Aggregates

Best Practices

Next Steps

Launch an Instance

Server Groups

Compute User Guide

Compute Admin Guide

Core Services

Other Services

​Overview

​How Availability Zones Work

​List Available Zones

Open the Launch Instance dialog

View available zones

Authenticate

List all zones

​Launch an Instance into a Specific Zone

Open the Launch Instance dialog

Select the availability zone

Complete instance configuration

​Relationship to Host Aggregates

​Best Practices

​Next Steps

Launch an Instance

Server Groups

Compute User Guide

Compute Admin Guide

Overview

How Availability Zones Work

List Available Zones

Launch an Instance into a Specific Zone

Relationship to Host Aggregates

Best Practices

Next Steps