Optimize your Testing Pipeline: Reduce Upload Times and Debug Faster

HyperExecute: A Solution to Accelerate Your Testing Productivity

HyperExecute enhances testing efficiency with features like Incremental Upload and FailFast, minimizing waiting times and accelerating feedback loops. Let's see how:

Incremental Upload: No More Uploading the Entire Codebase Every Time

Our customer, a fast-growing startup, relied on the Jenkins CI pipeline for their automated testing. Every time they encountered a failed test or made even a tiny code tweak, such as fixing a minor bug, it triggered their complete testing cycle.

This meant the entire codebase, a massive collection of files, had to be fetched onto Jenkins and then get triggered. This resulted in long waiting times for their QA team, eventually hindering and slowing down the entire process of diagnosing and fixing issues.

This drastic reduction in the test payload has several key benefits:

• Reduced Waiting Time: Developers no longer have to wait for lengthy uploads, allowing them to push code changes and iterate faster.
• Faster Feedback Loop: The QA team receives test results quicker, enabling them to identify and address issues promptly.
• Improved Efficiency: It frees up valuable time for both developers and testers, allowing them to focus on higher-level tasks.

differentialUpload ensures complete codebase testing while focusing upload efforts on the updated code portion. This created a win-win situation for the development and QA teams, leading to a more productive and efficient testing process.

You switch incremental upload on in the HyperExecute YAML file. The differentialUpload block takes enabled: true plus an optional ttlHours value that sets how long the cached codebase stays reusable, accepting 1 to 360 hours and defaulting to 60. A repository left untouched for longer than that TTL uploads in full on its next run, so teams on slow-moving branches should raise the value rather than assume the cache is still warm.

Faster Feedback, Faster Fixes: Eliminate the Waiting Time with Failfast

With the Jenkins-based CI pipeline, the QA team encountered another hurdle. Whenever a test failed, they had to rerun the entire test suite. This could take several hours, depending on the suite's complexity.

This meant that even if a critical bug was causing early failures, the QA team had to wait for the entire test suite to finish before they could debug the root cause of the problem. This was causing significant delays in the testing process, leading to frustration and decreased productivity among team members. HyperExecute's FailFast feature automatically aborted their test execution after a predefined number of consecutive failures. This helped their developers, or QA's, focus issues first and debug tests quickly.

With more immediate identification of failing tests, developers were able to receive feedback and begin fixing issues sooner. This led to a faster feedback loop between the development and QA cycles, accelerating their overall delivery process.

FailFast is set in the same YAML file. The failFast block takes maxNumberOfTests, the number of consecutive failures that aborts the job, and the counter resets as soon as one test passes, so scattered flaky failures do not halt a healthy run. Adding level: scenario counts failures per scenario instead of per test, and the documentation advises picking one level rather than applying both at once.