Designing for Resiliency (a.k.a. failover)

In many situations, customers need to change the server they are reporting on. Sometimes this can be planned, sometimes it can be spontaneous. This article covers how to fail-over an Engine to a new server.

How Windward Engine Fail-Over Works

Fail-over works very simple with Windward: simply stop reporting on your old Engine server, and start reporting on your new Engine server. 

What is the Upper Limit of Supported Fail-Over

Fail-over is lenient with Windward to cover many different cases. To look at the most extreme case of fail-over allowed, where a user needs to fail-over ever reporting server they have licensed simultaneously:

  1. Users are able to fail-over every Engine server they have licensed simultaneously.
  2. Users can overlap outputting reports from the old Engine server and the new Engine server while the fail-over takes place for any single server.
  3. Users must stop reporting from the old Engine server within a reasonable amount of time of performing the fail-over for any single server.

For this most extreme case, at one given time the user can at maximum be outputting from every old instance of Engine server and every new instance of Engine server simultaneously.

Limitations of to Fail-Over

Users are not allowed to continue reporting on the old Engine server for an unreasonable amount of time after performing fail-over to a new server. Users who continue to output on both servers for an unreasonable amount of time will have their case assessed on an individual basis, and may have all fail-over capabilities removed from their account.

When conducting fail-over, please avoid starting any new output on your old server to avoid any violations of Windward fail-over.


Add your comment

E-Mail me when someone replies to this comment

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.