Aftercare

~ by Randy Wagner

Going Live with a Policy Administration System (PAS) is an Exciting and a Crazy Time

Going live with a Policy Administration System (PAS) is very exciting, often the result of months or years of work and millions of dollars spent. The fear of going live can be paralyzing to product owners who see all the foibles lingering in the system. However, once the final “Go” decision is made, there is a moment of calm. Then customers hit the system and it can get crazy, especially for testers.

Managing Deferred Functionality and Defects Alongside New Production Defects is Tricky

No system goes live without some deferred defects or functionality. That’s the nature of software development. For organizations transitioning to a new system, it gets complex very quickly. Clean up of those deferred items takes on a new urgency. Then there is the immediate production support to handle defects that may occur in production. Still more, even if the full suite of applications is deployed, there are likely other lines of business or states to follow. Keeping track of multiple paths with typically limited resources can be an enormous challenge.

How to Keep Track of Multiple Paths with Limited Resources

There are ways of managing the chaos – Automate, automate, automate.

Automate and Optimize Regression Testing Cycles

QA never has enough resources to do the level of regression testing needed across multiple code lines. Using tools like GFIT, Katalon, and others is a force multiplier that’s indispensable these days.

  1. Identify the 80% bulk of transactions, automate those, and execute at multiple levels.
  2. Make the scripts available to Dev so it can be used to validate in local environments.
  3. Add them to the deployment tools so that smoke testing is fast and accurate.

Shifting left is the paradigm to catch defects even before they get to QA.

Use Scripts in Production Systems and Allocate QA Personnel Efficiently

For a system that’s in production, the scripts will be working against stable code where they can be of the most value.

  1. Use them in hot fix situations where time is critical and regression testing is always minimized.
  2. Add them to the Production Support stream to ensure new changes don’t break existing functionality.
  3. Use them in the next phase for the same reason.

Allocate the Testing Group Against the Multiple Streams

Experience and abilities vary between testers but it is better to align average testers to Production support where the issues are typically very well defined. Use the heavy hitters against new functionality that requires more planning, innovative thinking, and diligent execution.

Discipline is Key to a Successful Post Go Live Scenario

Code management is crucial. Typically, this happens outside the QA group, but QA is most affected by it.

Clear and precise packaging and deployment of code prevents the wrong items from bleeding across streams. QA typically gets the short end of this stick and needs to catch the issue quickly before it compounds with successive code changes. Development should be catching these issues during unit testing in the common Dev environment but that’s rarely the case.

Automation and smoke testing are the necessary bulwarks to flag these inconsistencies. Post Go Live is very exciting but it is packed with coordination and testing. Plan your teams in advance, develop your automation in advance, and regression test everything you can. It will all settle down to a more reasonable tempo soon enough.


Randy Wagner is Director of Quality Assurance for CastleBay Companies.  He has 20 years of consulting experience across private and public sectors, Guidewire InsuranceSuite, InsuranceNow, and Duck Creek, with specializations in quality assurance, configuration management, and automation.