Knapsack Pro wraps your current test runner(s) and works with your existing CI infrastructure to parallelize tests optimally:
- Dynamically splits your tests based on up-to-date test execution data
- Is designed from the ground up for CI and supports all of them
- Tracks your CI builds to detect bottlenecks
- Does not have access to your source code and collects minimal test data (with opt-in encryption)
- Enables you to export historical metrics about your CI builds
- Supports out-of-the-box any Ruby test runners, Cypress, Jest (and provides both SDK and API to integrate with any other language)
- Replaces local dependencies like Redis with an API and runs your tests regardless of network problems
Knapsack Pro API
The Knapsack Pro API splits test files between parallel CI nodes based on the test execution time of your previous CI runs. See Test Orchestration Modes below for more information.
Knapsack Pro Client
The Knapsack Pro Client is a library to install in your project that wraps your current test runners/frameworks. It connects with the Knapsack Pro API to:
- Retrieve the list of tests to run for the current parallel CI node
- Record the test execution time, including additional information from the CI environment variables:
- Commit hash
- Branch name
- Number of parallel CI nodes
- CI node index
Knapsack Pro comes with a web UI that displays graphs and stats about your test runs:
Tests Orchestration Modes
Queue Mode (Dynamic Split)
Knapsack Pro splits your tests dynamically (Queue Mode) between parallel CI nodes.
Each Knapsack Pro Client (parallel CI node):
- Retrieves from the Knapsack Pro API a subset of tests to execute
- Runs those tests
- Repeats 1. and 2. as long as there are more tests in the API queue
- Sends the test execution times to the API for the subsequent CI runs
With Queue Mode, Knapsack Pro distributes tests so that each CI node finishes at the same time. Thanks to that, your CI build time is as fast as possible.
Knapsack Pro allows distributing the tests in a static way (Regular Mode) too, but it’s rarely a good solution:
- Tests themselves, especially end to end, have varying execution times
- Failing tests execute faster than green ones
- CI nodes may differ in hardware/performance
- CI nodes boot at different times
- CI nodes are more or less busy when the environment is shared/virtualized
- CI nodes may load their caches at dissimilar times
- Test execution starts at different times on different CI nodes
- Flaky tests are.. Well, flaky!
Regular Mode (Static Split)
Even if Queue Mode is recommended, Knapsack Pro supports a static split of tests between parallel CI nodes (Regular Mode).
In this mode, the Knapsack Pro API returns to each Knapsack Pro client the subset of tests to execute in one batch.
However, as explained in the Queue Mode section, this will result in a suboptimal split.
In case of network issues, Knapsack Pro logs a warning and automatically switches to Fallback Mode to run your tests with a split based on test file names.
Split by test examples
Knapsack Pro can parallelize tests across CI nodes by example. This is useful when you have slow test files but don't want to manually split test examples into smaller test files.
Spot CI nodes
When you use spot CI nodes like Google Cloud Preemptible VMs or Amazon EC2 Spot Instances, Knapsack Pro gracefully handles preemptions: when you retry a node that was shut down the same tests are executed.
Retry only failed CI nodes
Knapsack Pro remembers what tests each CI node runs, so you can safely retry single nodes instead of re-running the entire suite. In some cases, you can retry only the tests that failed!
SAML-based single sign-on (SSO)
Besides traditional authentication (magic links, email/password), Knapsack Pro supports your organization's Identity Provider (IdP) like Okta, OneLogin, Google Workspace, or any other SAML SSO provider.
If you have questions, feel free to contact us directly and we will be happy to help.