Availability Zones in vSphere

Page last updated:

This topic describes availability zones (AZs) in vSphere and an example of how an operator might scale an app across AZs.

Overview

Pivotal defines an AZ as an operator-assigned, functionally-independent segment of network infrastructure. In cases of partial infrastructure failure, a tile distributes and balances all instances of running apps across remaining AZs. Strategic use of Availability Zones contributes to the fault tolerance and high availability of a tile deployment.

vSphere supports distributing deployments across multiple AZs. For more information, see Step 4: Create Availability Zones in Configuring BOSH Director on vSphere.

It is recommended that customers use three Availability Zones to operate a highly available tile deployment.

Balancing Across AZs During Failure: Example Scenario

An operator scales an app to four instances in a tile environment distributed across three availability zones: A1, A2, and A3. The environment allocates the instances according to the Diego Auction.

Az er 1

If A1 experiences a power outage or hardware failure, the two app instances running in A1 terminate while the app instances in zones A2 and A3 continue to run:

Az er 2

If A1 remains unavailable, the tile balances new instances of the app across the remaining availability zones:

Az er 3

Create a pull request or raise an issue on the source for this page in GitHub