Accelerating CI with AWS CodeBuild: Parallel take a look at execution now out there

I’m excited to announce that AWS CodeBuild now helps parallel take a look at execution, so you may run your take a look at suites concurrently and scale back construct instances considerably.

With the demo challenge I wrote for this put up, the entire take a look at time went down from 35 minutes to six minutes, together with the time to provision the environments. These two screenshots from the AWS Management Console present the distinction.

Sequential execution of the take a look at suite

Parallel execution of the take a look at suite

Very lengthy take a look at instances pose a big problem when operating continuous integration (CI) at scale. As initiatives develop in complexity and workforce measurement, the time required to execute complete take a look at suites can improve dramatically, resulting in prolonged pipeline execution instances. This not solely delays the supply of latest options and bug fixes, but in addition hampers developer productiveness by forcing them to attend for construct outcomes earlier than continuing with their duties. I’ve skilled pipelines that took as much as 60 minutes to run, solely to fail on the final step, requiring an entire rerun and additional delays. These prolonged cycles can erode developer belief within the CI course of, contribute to frustration, and finally decelerate the whole software program supply cycle. Moreover, long-running exams can result in useful resource rivalry, elevated prices due to wasted computing energy, and decreased general effectivity of the event course of.

With parallel take a look at execution in CodeBuild, now you can run your exams concurrently throughout a number of construct compute environments. This function implements a sharding method the place every construct node independently executes a subset of your take a look at suite. CodeBuild offers surroundings variables that determine the present node quantity and the entire variety of nodes, that are used to find out which exams every node ought to run. There is not any management construct node or coordination between nodes at construct time—every node operates independently to execute its assigned portion of your exams.

To allow take a look at splitting, configure the batch fanout part in your buildspec.xml, specifying the specified parallelism degree and different related parameters. Additionally, use the codebuild-tests-run utility in your construct step, together with the suitable take a look at instructions and the chosen splitting methodology.

The exams are break up primarily based on the sharding technique you specify. codebuild-tests-run presents two sharding methods:

• Equal-distribution. This technique types take a look at recordsdata alphabetically and distributes them in chunks equally throughout parallel take a look at environments. Changes within the names or amount of take a look at recordsdata may reassign recordsdata throughout shards.
• Stability. This technique fixes the distribution of exams throughout shards by utilizing a constant hashing algorithm. It maintains current file-to-shard assignments when new recordsdata are added or eliminated.

CodeBuild helps computerized merging of take a look at studies when operating exams in parallel. With computerized take a look at report merging, CodeBuild consolidates exams studies right into a single take a look at abstract, simplifying end result evaluation. The merged report contains aggregated cross/fail statuses, take a look at durations, and failure particulars, decreasing the necessity for guide report processing. You can view the merged ends in the CodeBuild console, retrieve them utilizing the AWS Command Line Interface (AWS CLI), or combine them with different reporting instruments to streamline take a look at evaluation.

Let’s take a look at the way it works
Let me reveal easy methods to implement parallel testing in a challenge. For this demo, I created a really fundamental Python challenge with a whole lot of exams. To velocity issues up, I requested Amazon Q Developer on the command line to create a challenge and 1,800 take a look at circumstances. Each take a look at case is in a separate file and takes one second to finish. Running all exams in a sequence requires half-hour, excluding the time to provision the surroundings.

In this demo, I run the take a look at suite on ten compute environments in parallel and measure how lengthy it takes to run the suite.

To achieve this, I added a buildspec.yml file to my challenge.

model: 0.2

batch:
  fast-fail: false
  build-fanout:
    parallelism: 10 # ten runtime environments 
    ignore-failure: false

phases:
  set up:
    instructions:
      - echo 'Installing Python dependencies'
      - dnf set up -y python3 python3-pip
      - pip3 set up --upgrade pip
      - pip3 set up pytest
  construct:
    instructions:
      - echo 'Running Python Tests'
      - |
         codebuild-tests-run 
          --test-command 'python -m pytest --junitxml=report/test_report.xml' 
          --files-search "codebuild-glob-search 'exams/test_*.py'" 
          --sharding-strategy 'equal-distribution'
  post_build:
    instructions:
      - echo "Test execution accomplished"

studies:
  pytest_reports:
    recordsdata:
      - "*.xml"
    base-directory: "report"
    file-format: JUNITXML 

There are three components to focus on within the YAML file.

First, there’s a build-fanout part beneath batch. The parallelism command tells CodeBuild what number of take a look at environments to run in parallel. The ignore-failure command signifies if failure in any of the fanout construct duties could be ignored.

Second, I take advantage of the pre-installed codebuild-tests-run command to run my exams.

This command receives the whole record of take a look at recordsdata and decides which of the exams should be run on the present node.

• Use the sharding-strategy argument to decide on between equally distributed or steady distribution, as I defined earlier.
• Use the files-search argument to cross all of the recordsdata which can be candidates for a run. We advocate to make use of the offered codebuild-glob-search command for efficiency causes, however any file search software, comparable to discover(1), will work.
• I cross the precise take a look at command to run on the shard with the test-command argument.

Lastly, the studies part instructs CodeBuild to gather and merge the take a look at studies on every node.

Then, I open the CodeBuild console to create a challenge and a batch construct configuration for this challenge. There’s nothing new right here, so I’ll spare you the main points. The documentation has all the main points to get you began.  Parallel testing works on batch builds. Make positive to configure your challenge to run in batch.

Now, I’m able to set off an execution of the take a look at suite. I can commit new code on my GitHub repository or set off the construct within the console.

After a couple of minutes, I see a standing report of the completely different steps of the construct; with a standing for every take a look at surroundings or shard.

When the take a look at is full, I choose the Reports tab to entry the merged take a look at studies.

The Reports part aggregates all take a look at information from all shards and retains the historical past for all builds. I choose my most up-to-date construct within the Report historical past part to entry the detailed report.

As anticipated, I can see the aggregated and the person standing for every of my 1,800 take a look at circumstances. In this demo, they’re all passing, and the report is inexperienced.

The 1,800 exams of the demo challenge take one second every to finish. When I run this take a look at suite sequentially, it took 35 minutes to finish. When I run the take a look at suite in parallel on ten compute environments, it took 6 minutes to finish, together with the time to provision the environments. The parallel run took 17.9 % of the time of the sequential run. Actual numbers will range along with your initiatives.

Additional issues to know
This new functionality is suitable with all testing frameworks. The documentation contains examples for Django, Elixir, Go, Java (Maven), Javascript (Jest), Kotlin, PHPUnit, Pytest, Ruby (Cucumber), and Ruby (RSpec).

For take a look at frameworks that don’t settle for space-separated lists, the codebuild-tests-run CLI offers a versatile various by means of the CODEBUILD_CURRENT_SHARD_FILES surroundings variable. This variable incorporates a newline-separated record of take a look at file paths for the present construct shard. You can use it to adapt to completely different take a look at framework necessities and format take a look at file names.

You can additional customise how exams are break up throughout environments by writing your personal sharding script and utilizing the CODEBUILD_BATCH_BUILD_IDENTIFIER surroundings variable, which is routinely set in every construct. You can use this system to implement framework-specific parallelization or optimization.

Pricing and availability
With parallel take a look at execution, now you can full your take a look at suites in a fraction of the time beforehand required, accelerating your growth cycle and enhancing your workforce’s productiveness.

Parallel take a look at execution is offered on all three compute modes provided by CodeBuild: on-demand, reserved capability, and AWS Lambda compute.

This functionality is offered at the moment in all AWS Regions the place CodeBuild is obtainable, with no further value past the usual CodeBuild pricing for the compute sources used.

I invite you to strive parallel take a look at execution in CodeBuild at the moment. Visit the AWS CodeBuild documentation to study extra and get began with parallelizing your exams.

— seb

PS: Here’s the immediate I used to create the demo utility and its take a look at suite: “I’m writing a blog post to announce codebuild parallel testing. Write a very simple python app that has hundreds of tests, each test in a separate test file. Each test takes one second to complete.”

How is the News Blog doing? Take this 1 minute survey!

(This survey is hosted by an exterior firm. AWS handles your data as described within the AWS Privacy Notice. AWS will personal the information gathered by way of this survey and won’t share the data collected with survey respondents.)