Fail Fast — Your Fastest Path to Success — DigyKube BYOC Can Help

Digy4Tech
4 min readMay 28, 2022

Fail Often — Your Fastest Path To Success — How DigyKube BYOC Can Help To Get There.

Your success can involve your failures — as ironic as it may sound. Read on to understand the value of the statement.

In case you are wondering what Fail fast is — it refers to running tests as soon as a code is checked in and identifying the failure points fast. You cannot fail fast if you cannot run fast.

In software, life cycle testing often happens towards the end of the lifecycle. Everyone is familiar with the situation where every other timeline in the delivery cycle is pushed out and finally, when it comes to testing we often hear how the release dates can’t be moved. The product is already developed and the business team wants to push it out to the market as soon as possible. Testing has to find magical ways to finish up fast and deliver with quality at the same time. All the estimates go out the window and it’s all hands-on deck.

Some of the common challenges testing runs into are:

1. Development is delayed and there is not enough time to complete the planned testing. The release date can’t be moved so what can we test with the limited time left?

2. Tester finds a showstopper close to the release. Why did we find the issue so late?

3. Performance testing team finds issues ,even with a small number of users close to the release and can’t scale to high volume tests. The fix needs major changes and we have to plan the fix for the next release.

4. Development team doesn’t have the necessary infrastructure to test. We did some basic testing and left the remaining for the QA team to cover during their testing.

5. We have a high percentage of automation coverage but it still takes hours or even days to complete the automation run.

The list goes on…

We all live through these challenges in every organization but how do we overcome them?

Let us look at some realistic approaches to address these challenges.

Automated Testing Pipeline — Access to All: Allow to Fail on Demand

1. Build an automated testing pipeline that all teams in your organization can run and get the feedback themselves. Fail fast and fix issues earlier in the cycle. Fixing issues earlier in the cycle is always more cost-effective than finding and fixing issues later in the cycle. Now you don’t have to limit your automation pipeline to the QA team only due to limited scalability in SaaS browserfarms. DigyKube BYOC allows you run hundreds of browsers in parallel in the cloud and everyone in the organization can run tests and get instant feedback. Developers will be able to deliver quality code and testers will have fewer issues to deal with. Quality doesn’t have to be an afterthought. Quality can be a Day1 practice in software delivery. We shouldn’t be waiting for integration testing to start before we run our first test.

Run critical tests frequently to fail fast

2. Don’t shoot the messenger. Perhaps it’s time to rethink the testing strategy. If you fail fast and keep testing close to the code for your critical scenarios, your risk of finding late surprises or critical bugs will come down significantly. DigyKube BYOC increases your capacity to run critical tests more frequently than ever before and reduce the risk of late surprises. QA teams won’t have keep justifying why the issue couldn’t be found earlier in the cycle.

Fail performance earlier in the cycle

3. Performance testing can start much earlier in the testing cycle. Starting earlier will lower the number of performance tests and tools that allow performance tests to happen earlier and closer to the development stage. Consider creating a Performance testing pipeline of low-volume performance testing using DigyKube BYOC as load testing infrastructure so you can start receiving feedback earlier.

Extend Test Infrastructure to Dev to test and fail fast

4. Often the development team doesn’t have the necessary infrastructure that they need to properly test their check-ins. Choose tools in your testing cycles that can scale to support development teams as well. They may need browser farms to run multiple browser-based tests. Adopt cost-optimal and scalable browser farm solutions that allow you to extend to development teams. Otherwise, even if the Development team wants to test, they will not be able to and will depend on integrated testing to give them feedback. Testing will remain an afterthought in the software cycle.

Scalable Infrastructure to Support Fail Fast

5. Having high automation coverage is not enough. When you are building up a large volume of automation, consider building up a scalable test infrastructure that you can run your test in parallel.

You will still be struggling to bring the acceleration into testing if you are running tests in sequential mode, local execution or with limited parallel execution capability. Adopt DigyKube BYOC that scales on-demand and scales down when not needed to bring the acceleration.

Automation without a scalable infrastructure is like running a race car on a city road. Change your road to experience the true speed of your car.

Run tests earlier and fail fast to fix them earlier. Bring shift left to your testing to accelerate and minimize your risk. Build quality with your different mindset — test close to the code. Know your testing goal before you build every part of your testing tool strategy. Shift left is possible and knowing there are options to achieve it will allow you to act differently. Know Before You Act!

--

--