Integrating with GitLab CI
Keploy can integrated with GitLab CI to streamline your testing process and ensure continuous testing as part of your CI/CD pipeline.
Create pipeline
To integrate the Keploy in GitLab, we first need to install and setup by adding the following steps to our .gitlab-ci.yml : -
---
stages:
- test
keploy-test-job: # This job runs in the test stage.
image: ubuntu:22.04
stage: test
before_script:
## Add the dependencies
- apt-get update && apt-get install -y curl python3 python3-pip git kmod linux-headers-generic bpfcc-tools sudo
- git clone https://github.com/keploy/samples-python
- cd flask-mongo
- mkdir -p /sys/kernel/debug
- mkdir -p /sys/kernel/tracing
script:
## Steps to install Keploy and run application
...
Note: if you are using
arm_64as runner use below to download keploy binary
curl --silent --location "https://github.com/keploy/keploy/releases/latest/download/keploy_linux_arm64.tar.gz" | tar xz --overwrite -C /tmp
Now that we have Keploy installed, and all ready, we need switch to path where keploy folder is present in our application and install all the application related dependencies. Since we are using flask-mongo sample-application, steps in our script: would look like below:-
script:
## Steps to install Keploy and run application
# Install Keploy
- curl --silent -O -L https://keploy.io/install.sh && source install.sh
# Mount required filesystems
- mount -t debugfs nodev /sys/kernel/debug || true
- mount -t tracefs nodev /sys/kernel/tracing || true
# Run Keploy tests
- pip3 install -r requirements.txt
- keploy test -c "python3 app.py" --delay 50
In your .gitlab-ci.yml file, in last step we have keploy test command to run your keploy generated test suite, this sets up Keploy to replay the interactions it has generated and perform CI Testing.
📝 Note
Did you notice some weird stuff in the pipeline? Like kmod, linux-headers, /sys/kernel/debug...and thought, "Wait, am I hacking the kernel or something?" 😅
Don’t worry — these are just there because Keploy uses eBPF (a cool Linux feature) to trace your app’s behavior.
So we install kmod, linux-headers-generic, and bpfcc-tools to make that tracing possible.
Some CI systems don’t have /sys/kernel/debug and /sys/kernel/tracing by default, so we create them and mount debugfs and tracefs — it’s like giving Keploy the backstage pass it needs to watch your app in action.
No black magic. Just some low-level Linux stuff helping your tests run like magic! 🪄✨
We will get to see output : -
$ keploy test -c "python3 app.py" --delay 50
▓██▓▄
▓▓▓▓██▓█▓▄
████████▓▒
▀▓▓███▄ ▄▄ ▄ ▌
▄▌▌▓▓████▄ ██ ▓█▀ ▄▌▀▄ ▓▓▌▄ ▓█ ▄▌▓▓▌▄ ▌▌ ▓
▓█████████▌▓▓ ██▓█▄ ▓█▄▓▓ ▐█▌ ██ ▓█ █▌ ██ █▌ █▓
▓▓▓▓▀▀▀▀▓▓▓▓▓▓▌ ██ █▓ ▓▌▄▄ ▐█▓▄▓█▀ █▓█ ▀█▄▄█▀ █▓█
▓▌ ▐█▌ █▌
▓
version: 2.5.1
🐰 Keploy: 2025-04-14T16:22:12Z WARN python language detected. please use --language to manually set the language if needed
🐰 Keploy: 2025-04-14T16:22:13Z INFO keploy initialized and probes added to the kernel.
🐰 Keploy: 2025-04-14T16:22:14Z INFO starting UDP DNS server at addr :26789
🐰 Keploy: 2025-04-14T16:22:14Z INFO Keploy has taken control of the DNS resolution mechanism, your application may misbehave if you have provided wrong domain name in your application code.
🐰 Keploy: 2025-04-14T16:22:14Z INFO Proxy started at port:16789
🐰 Keploy: 2025-04-14T16:22:14Z INFO running {"test-set": "[test-set-0]"}
🐰 Keploy: 2025-04-14T16:22:14Z INFO starting TCP DNS server at addr :26789
* Serving Flask app 'app'
* Debug mode: on
WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
* Running on all addresses (0.0.0.0)
* Running on http://127.0.0.1:6000
* Running on http://127.0.0.1:6000
Press CTRL+C to quit
* Restarting with stat
* Debugger is active!
* Debugger PIN: 100-754-563
🐰 Keploy: 2025-04-14T16:23:04Z INFO starting test for of {"test case": "[test-1]", "test set": "[test-set-0]"}
127.0.0.1 - - [14/Apr/2025 16:23:04] "POST /students HTTP/1.1" 200 -
Testrun passed for testcase with id: "test-1"
--------------------------------------------------------------------
🐰 Keploy: 2025-04-14T16:23:04Z INFO result {"testcase id": "[test-1]", "testset id": "[test-set-0]", "passed": "[true]"}
<=========================================>
TESTRUN SUMMARY. For test-set: "test-set-0"
Total tests: 1
Total test passed: 1
Total test failed: 0
Time Taken: "50.01 s"
<=========================================>
COMPLETE TESTRUN SUMMARY.
Total tests: 1
Total test passed: 1
Total test failed: 0
Total time taken: "50.01 s"
Test Suite Name Total Test Passed Failed Time Taken
"test-set-0" 1 1 0 "50.01 s"
<=========================================>
🐰 Keploy: 2025-04-14T16:23:05Z INFO calculating coverage for the test run and inserting it into the report
Wrote JSON report to coverage.json
🐰 Keploy: 2025-04-14T16:23:05Z INFO [Total Coverage Percentage: 64%]
🐰 Keploy: 2025-04-14T16:23:05Z INFO stopping Keploy {"reason": "replay completed successfully"}
🐰 Keploy: 2025-04-14T16:23:05Z INFO proxy stopped...
🐰 Keploy: 2025-04-14T16:23:05Z INFO eBPF resources released successfully...
Cleaning up project directory and file based variables
00:00
Job succeeded
And... voila! You have successfully integrated keploy in GitLab CI/CD pipeline 🌟
Integrating Keploy with GitLab CI automates the testing process, ensuring that tests are run with every commit and merge request. And by running tests automatically in CI pipeline, you can catch issues early and ensure that your application remains stable and reliable.
📦 Need the Full Pipeline?
If you’re thinking, “This pipeline looks cool, but I need the whole thing to integrate with your application!” — well, you're in luck! Check it out here and get ready to copy-paste your way to success! ✨🚀
Hope this helps you out, if you still have any questions, reach out to us .
Contact Us
If you have any questions or need help, please feel free to reach out to us at hello@keploy.io or reach out us on
