Testing a module before putting it into production
Installing a module straight onto the store that sells is the most avoidable cause of weekend outages. A test copy changes everything — provided you know what it proves and what it doesn’t.
What a test copy genuinely shows
A faithful copy of the store, on a separate address, gives three things no product listing can: the module’s real effect on your data rather than on a demo catalogue, its cost in page generation time, and the freedom to do destructive things — disable, uninstall, start again — with no commercial consequence.
It is also the only place to check interaction with modules already installed. A module behaves differently alone and among fifteen others, and that difference is only observable on a copy of the site as it really is.
The three traps of a staging copy
The domain-locked licence. Many commercial modules check the domain name they run on. On a test address the module may refuse to activate, or activate in a degraded mode. Check first whether the licence covers a test installation: some publishers plan for it explicitly, others don’t.
Production data copied as is. A copy contains real addresses, real email addresses and real orders. Outgoing email must be cut and live payments disabled before the first action, otherwise a test order sends a real message to a real customer.
Environment differences. A module that works in test and not in production nearly always reveals a gap in PHP version, installed extensions or server configuration. A useful staging copy is an identical one, not an approximate one.
Related pages
Frequently asked questions
Should staging be permanent or one-off?
Does my module licence cover a test copy?
Can we test without copying the real data?
How long should the copy be kept after the work?
Describe your need in one minute
A few targeted questions so I can reply with an estimate rather than another questionnaire.