I’m attempting to gather research on why companies build test networks as a project that I’m working on. I’ve searched many, many Web sites and tech pubs and haven’t see many article providing some of the justifications and ROI statistics around this subject. Can you point out some resources for me?
I’m attempting to gather research on why companies build test networks as a project that I’m working on. I’ve searched many, many Web sites and tech pubs and haven’t see many article providing some of the justifications and ROI statistics around this subject. Can you point out some resources for me?
— Maurice Moore
I can do better than point you to articles or white papers for justifications or ROI data on building test networks, I can cite practical experience for you.
In a former life, I was the senior network engineer for an international bed manufacturer located in the US. The company at that time ran on an Oracle-based system for accounting, sales and shipping functions. It wasn’t unusual to have problems with Oracle patches not installing as Oracle support said they would – if they installed at all. Being able to apply the patches/updates on a test system allows for testing to be done and any debugging necessary without having to disable production access. This also allowed for new functionality to be tested without damaging production data or accidentally overloading the production database server with some misconfigured triggers.
The test network also had a secondary purpose. In the event that one or more parts of the production network went down for any reason and couldn’t be quickly brought back online, one or more components of the test network could be switched over to use on the production network in a matter of minutes. When working with additional third-party vendors who offered functionality add-ons to this type of system, you could have the first install done on the test system and when everything checked out as expected, you could proceed with the real install on your production system with the reasonable expectation that everything should work out as expected.
No, a test network will not be cheap, since you are essentially looking at building a redundant server configuration. But when you look at the benefits of a test network, the initial cost could pale beyond that of the time and data you could potentially lose during a planned upgrade or change gone awry. If you do some reading on Sarbanes-Oxley, this might help give you some additional ammunition for being able to proceed with a test network.
Continue the discussion. Jump into our forum on test networks.
