Testing Environment Management Resolutions for 2016

To follow up our previous post on environment management resolutions for 2016, here is a post focused on specific resolutions for Test Environment Management (TEM). Most of the challenge for TEM relates to unknowns surrounding the testing process: How long is a testing process going to take? What are the shifting data requirements for test environments? And, how can a Test Environment Manager plan ahead for the unexpected twists and turns of an unpredictable software development lifecycle.

Whether it is the constrained timelines of application development groups or the ability of your data management team to prepare appropriate data sets for testing, testing software properly requires the orchestration of several variables. Testing environments are often a source of delay in the enterprise, but it doesn’t have to be this way. In 2016, there are a few resolutions that can help you manage uncertainty and more effectively respond to the needs of the development teams you support:

Resolution #1. My environment schedules will allow for several test cycles.

What use is a testing environment in QA if you only have a single day to use it? If you work in an organization that shares testing environments you’ve seen what happens when teams don’t have enough time to properly test their software, and the results can be catastrophic for a business. When teams already under intense time pressures are only offered a testing environment for a single day or a few hours there is a tendency to ignore errors that would otherwise be addressed if only they had more time to test.

As an environment manager in charge of scheduling time for various projects you should adjust your environment booking schedules to allow adequate time for multiple test cycles. If you don’t allow time for teams to identify and address bugs in multiple QA cycles you’ll see the end-result in revenue loss and production outages. You test software for a reason, and you should give development teams time to address the unknown bugs testing is designed to uncovered. Use Plutora to create realistic schedules for your testing environments.

Resolution #2. I will automate test environment setup.

Teams sitting around waiting for weeks while DBAs manually prepare data sets is a massive waste of time, yet this is the norm across many industries. In this age of “DevOps” automation, any process that hasn’t yet been automated should have a solid plan for automation in 2016. Let 2016 be the year you decide to use tools like Puppet or Chef to automate the setup and teardown required for full automation of testing environments.

The end-result of more automation isn’t just time saved it will change the way Test Environment Managers think about environments as being more agile. This will pay for itself in terms of increased efficiency and new opportunities to rapidly test software. If it only takes 30 minutes or an hour to stand up a new environment with a new set of VMs capturing both application state then development lifecycles can be accelerated. Don’t delay. Automate test environments this year, and use Plutora to keep track of the effort involved in implementing automation.

Resolution #3. Performance environments will more closely resemble production.

Nothing is worse than a failure in production during a critical time of the year. Whether you are a retailer during Cyber Monday or an investment bank during a busy trading day you take steps to run capacity tests. You prove a system can scale to meet demand before it has to, or do you? If your performance environment doesn’t truly match production or if your testing data isn’t as large or as varied as real production data you will miss critical defects that only make themselves known at scale.

You can avoid these problems by properly sizing your performance environment to make sure that your data sets resemble production data. Make 2016 the year that you provision enough capacity to run real performance tests, and use this year as an opportunity to build performance test environments that can be relied upon to produce meaningful test data. Use the efficiency achieved with Plutora and the excess capacity you’ve identified with the tool it has provided to create larger performance testing clusters. Use the time saved to build better, more “accurate” environments for performance testing.

Resolution #4. I will measure our response to environment booking requests.

In many software development organizations teams have mastered the art of waiting, politely, for what can seem like an eternity for testing environments. Many teams build this delay into their schedules in anticipation that it will take weeks to provision the proper environments for testing.

If you’ve been tasked with responsibility for environment management use a tool like Plutora to help you keep track of your customer’s requests for test environments and use the data generated by Plutora to commit your teams to a continuous process of improvement. Don’t let environment management become yet another enterprise bottleneck. Instead, track your response time and set aggressive goals for reducing the time it takes to satisfy requests for test environments.

Resolution #5. I will identify underutilized testing resources.

What’s worse then telling a team that there are not enough resources for a testing environment? Telling them this when you know that there are several environment allocated to other sitting around unused… for months.

When you are responsible for test environment management, you are also responsible for measuring environment utilization. Make sure that teams understand that environments that are unused will be reclaimed, and (as I mentioned in the previous post) make sure to put an expiration date on a team’s lease for a test environment.

You can do just this with the help of Plutora. Think of Plutora as a system that allows teams to checkout and reserve environments on a temporary basis. With Plutora you have a dashboard that gives you insight into what’s happening with every environment in one place. Just having this view is enough to address the problem of environments sitting around unused and ignored.