Massive Duplicate Alerts Auto‑Created by Git Folders After Recent Databricks Update

[kcheng]

Hello Databricks Support Team,

I’m experiencing a severe issue in my Databricks workspace related to the new Git‑Integrated Alerts behavior. Overnight, my workspace went from 67 alerts to nearly 1,000 alerts, all of which appear to have been auto‑generated from Git folders without any action taken on my part. Those hundreds of duplicate alerts are all marked as 'Not Run' and Stopped, and has same name as alerts stored in Git repositories. This behavior did not exist as of Friday (March 21, 2026), but appeared suddenly the morning of March 23, 2026.

This issue has severely impacted our alert management workflow since we created single Git folder for task separation and there's no option in workspace UI to disable Git-source alert syncing. This is extremely unintuitive and disruptive, and appears to have been triggered by a backend update rather than user action.

I would like to ask for answer on following request:

• Please confirm whether a recent Databricks update triggered Git folder alert syncing in our workspace.
• Is there a feature flag or workspace configuration to disable Git‑integrated alert auto‑creation?
• Guidance on safely cleaning up the hundreds of duplicate alerts created.
• Would there be a feature 'Delete at Once' for user selected Alert?

This issue severely affects usability for anyone using multiple Git repos inside a single Databricks workspace and seems inconsistent with typical alert management workflows.

We would appreciate your assistance in identifying the cause, disabling the behavior, and removing the auto‑generated alerts.

[Ashwin_DSA · Accepted solution]

Hi @kcheng,

Thanks for sharing the details. This looks like behaviour that will need workspace‑specific investigation by Databricks Support, rather than something the community can reliably diagnose or fix.

Because it resulted in a sudden, large volume of auto‑created alerts, I’d strongly recommend raising a Support ticket with your workspace URL, region, and a rough time window when the alerts appeared. That will let the Support and engineering teams check backend logs, confirm whether this is expected behaviour, and advise on safe cleanup options.

If this answer resolves your question, could you mark it as “Accept as Solution”? That helps other users quickly find the correct fix.