Calculating downtime in simple redundant networks

I have a simple redundant network of path A and path B. Using RBDs, I show this as path A in parallel with path B. So I'd have to have a failure in path A and B at the same time before my availability would be affected. In reality however, a failure in path A would cause a 30 second outage (downtime) while the software re-routes traffic down path B. Any ideas how I account this in my RBDs and availability calculations?

    Requires Free Membership to View

    Login

As with all downtime, you calculate it from a user's perspective. Can the user get his job done during the outage? If not, then you must count the 30 seconds as downtime. As far as the user can tell, it is no less downtime than occurs during a server crash.

This was first published in September 2004

Join the conversationComment

Share
Comments

    Results

    Contribute to the conversation

    All fields are required. Comments will appear at the bottom of the article.
    Back to top