Exalate looping synchronizations in Jira Cloud forge app

Update timeline

1. investigating Jul 13, 2026, 10:15 PM UTC

   We've received multiple reports of syncs looping after update to the new Jira Cloud forge app. Our initial investigation leads us to believe this is happening due to the change in the Proxy User account. We are actively working on determining the root cause and applying a fix.

2. investigating Jul 13, 2026, 11:48 PM UTC

   We continue to investigate this issue with our engineers to determine the root cause.

3. identified Jul 14, 2026, 04:58 AM UTC

   The root cause has been identified and we are working on a fix. Please expect another update within an hour.

4. identified Jul 14, 2026, 07:05 AM UTC

   The fix is currently being validated to find the best approach forward. We will provide another update within the hour.

5. identified Jul 14, 2026, 08:11 AM UTC

   Unfortunately, the testing of the fix revealed an unexpected issue and we are now fine tuning the fix in order ensure that it incorporates all possible use cases. We will have an update for you in an hour.

6. identified Jul 14, 2026, 09:09 AM UTC

   The new fix has been created and initial tests are promising. We are deep in QA with the new fix and will provide an update within the hour about the results. The inconvenience is regretted.

7. identified Jul 14, 2026, 10:11 AM UTC

   We are still testing the fix and results are still good. We are close to finalizing it. Next update will be within the hour.

8. identified Jul 14, 2026, 11:39 AM UTC

   The fix is ready and we are working on the deployment approach with minimal impact for customers. It will be systematically rolled out to all clients. We will provide an update in the next hour about the deployment.

9. identified Jul 14, 2026, 12:51 PM UTC

   We have deployed the fix to a few select customers and validated results further. The mass rollout is being planned right now. We will update you within the next hour about progress.

10. identified Jul 14, 2026, 02:24 PM UTC

    We are still on course for the patch deployment and the release is going through additional QA validation now (required for a full production deployment). We expect to have the next update within an hour.

11. identified Jul 14, 2026, 04:11 PM UTC

    The QA and testing is taking slightly longer than expected, but we expect to start the deployment within the next hour. We will update you before starting this.

12. monitoring Jul 14, 2026, 05:41 PM UTC

    The deployment of the hotfix: Exalate Classic 5.35.4 - Jira Cloud, in underway (https://community.exalate.com/t/jira-cloud-version-5-35-4/8215). The deployment is being closely monitored for any anomalies. We will update you in an hour.

13. monitoring Jul 14, 2026, 06:55 PM UTC

    The deployment continues uninterrupted. We will update you in due course as soon as it is fully completed.

14. resolved Jul 14, 2026, 10:14 PM UTC

    The deployment was completed successfully and the issue reported stands fully resolved. A detailed RCA will be published in due course.

15. postmortem Jul 28, 2026, 08:00 PM UTC

    **Impact:** Some Classic Exalate for Jira Cloud connections experienced repeated, duplicated synchronisations following an infrastructure migration to Atlassian's Forge platform. On affected connections customers saw: * **Duplicate activity** — the same update, comment, or field change applied to an issue over and over as the sync looped. * **Reopened / resurrected issues** — issues that had been closed being reopened, and other status or field values being overwritten back and forth between the paired issues. * **Distorted dashboards and reports** driven by the volume of duplicate changes. * **Broken automations and workflow validations** — because the integration user's name \(and underlying ID\) changed, customer-side Jira Automation rules and workflow validators that referenced the "Exalate" user by name stopped behaving as configured. * **Stuck or failed syncs**, and in some cases a connection going down, requiring manual retries or recovery. * Increased **manual effort** for teams to unstick syncs and correct affected issues. Not all customers were affected — impact depended on connection configuration. The most exposed were local connections and issues that had previously been created or updated under the original "Exalate" identity and were then touched under the renamed one. **Summary:** As part of moving Classic Exalate for Jira Cloud to Atlassian's Forge platform, the name shown for Exalate's integration user inside Jira changed \(from "Exalate" to a longer variant\). Exalate used that name to recognise its own updates and avoid re-synchronising them. Once the name changed, the app no longer recognised its own writes on affected connections and re-processed them as if they were new external changes — each write triggering another sync, producing loops. The same identity change also broke customer automations and validations that referenced the "Exalate" user by name. **Resolution:** A patch \(5.35.4\) that recognises Exalate's own updates by a stable user identifier rather than the display name was validated on affected customer environments and deployed across the Jira Cloud fleet. Synchronisation returned to normal once deployment completed. Atlassian has since restored and confirmed the integration user's name as "Exalate", and it is now stable — repairing automations that reference it by name. Where loops created duplicate or reopened items, we are working with affected customers to identify and correct the affected issues. **Preventive measures:** Loop-prevention no longer depends on a changeable display name; we are adding an impact-review and notification step around platform-driven changes and app updates, and automated detection of synchronisation loops and database saturation so recovery does not depend on customer reports. We apologise for the disruption and remain committed to the reliability of the Exalate platform.