[Bug][Config UI] The Stop functionality in Config-UI is not functioning when the current pipeline is running for a longer time.

[ParthibanPRP]

Search before asking

• I had searched in the issues and found no similar issues.

What happened

I am unable to stop the pipeline which was running for more than 13 hours. I tried hitting the "STOP" CTA in Config-UI screen but to no anvil. Also, I do not have logs for the specified issue. Moreover, the same issue was posted by team member in slack channel on Aug 18.

Currently I am using v1.0.3-Beata5

What do you expect to happen

Pipeline should stop when we hit the stop button in the status tab

How to reproduce

1. Create a project and assign the connection (consisting of GitHub, Jira and SonarQube Data sources) and start the pipeline to trigger.
2. Allow the pipeline to trigger for more than 2 hours
3. Hit the Stop button available in the Status tab.
4. You would notice the pipeline still in the running condition.

Anything else

No response

Version

v1.0.3-beta5

Are you willing to submit PR?

• Yes I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct

[dosubot]

Hi @ParthibanPRP! I'm Dosu and I’m helping the incubator-devlake team.

This issue with the Stop button not working for long-running pipelines is known in the community and is often linked to backend unresponsiveness, resource constraints, or plugin bugs. The Stop action in the UI sends a DELETE request to the backend to terminate the pipeline, but if the backend is frozen or the running task doesn't check for cancellation, the stop won't take effect. Users have reported similar problems when pipelines run for hours or days, especially with large data sources or limited memory environments.

Workarounds that have helped others include increasing container memory (e.g., from 4GB to 6GB), isolating problematic repositories, enabling shallow cloning, and upgrading to newer beta versions (like v1.0.3-beta6) if plugin bugs are suspected. If the backend becomes unresponsive, sometimes only restarting the container or pod will abort the pipeline. Collecting backend logs and noting which data source or plugin is running when the stop fails can help diagnose the root cause. More details and community experiences are discussed in related issues and discussions, here, and here.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}    

[klesh]

Can you share the logs of the devlake container?