Question

I’ve been with my current company for about four months now and I’ve noticed how several of our RnD scopes/documents use the term “lifecycle testing.”

I’ve always thought that this term would mean the entire testing phase of a project, but the context of the term suggests that it instead is when the software is tested with “live” or “real” data in a staging environment as close to the production environment as possible.

This has led me to wonder if I have misunderstood the meaning of the phrase, in which case, can somebody explain what lifecycle testing is supposed to be or mean?

Answer 1

A lifecycle of software is it’s behaviour in the following situations:

• Startup. Does it load correctly? Is it fast at startup? (Depends on what kind of software)
• Mid-life. Does it use much memory? Does it clean up memory? Does it do what it’s ought to do?
• Exeting. Does it cleanup resources correctly? Does it closes everything down well?

Lifecycle testing is very important for server applications, where it’s especially focussed on “mid-life” (it’s not an official term btw). Server apps may never crash while doing something importantly, and if they do: they shouldn’t bring down the complete system.

The clue “lifetime” of being “live” or “real” isn’t much true, it’s more being “alive” than “live”.

For instance; I’ve build a Flash client-application which is a “billboard” application, displayed at a large screen, and I am lifecycle-testing it:

• Graphics, does everything show up well? Not just the first minutes, but even 12 hours without restarting the app.
• Auto-update, does that work?
• etc.