Affected components
Update timeline
- identified Feb 27, 2026, 04:06 PM UTC
Since approximately 12:30 UTC a subset of calendar events failed to be synced to users' calendars. We've identified the source of the issue and released a fix, we are now assessing the impact and actions necessary to resolve the issue for previously impacted calendar events.
- identified Feb 27, 2026, 04:28 PM UTC
We have identified that affected events were those being written to Microsoft calendars and using the `integrated` conferencing option. The system is already retrying these failures and successfully writing them. We are looking to accelerate this process.
- identified Feb 27, 2026, 05:00 PM UTC
At approximately 12:30 UTC a change was released that prevented events being created in Microsoft calendars if integrated conferencing was being used. A fix was released and the last error observed at approximately 16:05 UTC. Since that time over 90% of the impacted events have been successfully synced due to standard retry mechanisms with the platform. We are working to enqueue additional attempts for all affected calendars to speed up this process and to be certain nothing has been missed.
- resolved Feb 27, 2026, 05:35 PM UTC
At approximately 12:30 UTC a change was released that prevented events being created in Microsoft calendars if integrated conferencing was being used. A fix was released and the last error observed at approximately 16:05 UTC. Over 90% of the impacted events were successfully synced due to standard retry mechanisms with the platform by 17:00 UTC, we also ran a sweep to proactively retry all remaining events which completed by 17:30 UTC. This incident is now resolved and operations are back to normal. We will be conducting a postmortem and will publish the findings of that next week.
- postmortem Mar 03, 2026, 03:12 PM UTC
On Friday February 27th 2026 at 12:28 UTC, a change was released to our production environment that caused a bug preventing events using Microsoft Integrated conferencing from being synced to calendars. At 16:05 UTC, a change was released to fix the issue and full functionality was restored. By 17:35 UTC, all previously impacted events had been successfully retried and synced and the incident was closed. In total, we identified 2,254 events that had been affected by this bug, all of which have now been successfully synced. We always ask the questions: * Could the issue have been identified sooner? * Could the issue have been resolved sooner? * Could the issue have been prevented? In this instance, the issue could and should have been identified by us earlier. The change that introduced this bug was released at 12:28 UTC, but we only became aware of the issue when a customer reported issues to us at 15:27 UTC. We never want to rely on customers having to tell us about issues in our platform before we are aware, so this is an instance where gaps in our system monitoring meant the issue went undetected for too long. We deployed a fix for this issue at 16:05 UTC, which was 37 minutes after the bug report email was sent, and 13 minutes after we identified the extent of the issue and opened an incident at 15:52 UTC. Therefore, we don’t believe the bug could’ve been resolved any sooner, as we acted quickly to release a fix once we were fully aware of the problem. Along with a gap in our alerting, the bug should have been prevented through automated testing, which would’ve flagged the issue before the initial change was deployed at 12:28 UTC. In this instance, our testing failed to check the logic down the path that Microsoft Teams integrated conferencing followed, because a check had been mocked instead of being integration tested, leading to a blind spot. We have addressed this and are reviewing where we have used test mocks which could hide integration issues. Actions: * Further guidelines around coding principles to be provided to our Engineering team regarding our approach to automated testing * Improve our system monitoring to provide specific alerting around this particular area
Looking to track Cronofy downtime and outages?
Pingoru polls Cronofy's status page every 5 minutes and alerts you the moment it reports an issue — before your customers do.
- Real-time alerts when Cronofy reports an incident
- Email, Slack, Discord, Microsoft Teams, and webhook notifications
- Track Cronofy alongside 5,000+ providers in one dashboard
- Component-level filtering
- Notification groups + maintenance calendar
5 free monitors · No credit card required