Microsoft Fabric Outage History

Microsoft Fabric experienced dataflow refresh issues, primarily in the North Europe region, leading to degraded performance for Dataflow Gen2 and related Fabric services. The incident began on Feb 26, 2026 5:19 PM UTC as Fabric indicated a degraded state across multiple regions, with Europe specifically reporting problems with Dataflow Gen2 refreshes. Users across regions reported slow Fabric operations and difficulties saving reports, including observations from the United States and Colombia, suggesting a broader but regionally concentrated impact. The outage lasted approximately 19 hours and 35 minutes, with the Europe Fabric component returning to normal by Feb 27, 2026 12:54 PM UTC according to status updates. Four status updates were issued during the incident to capture ongoing investigations and eventual restoration.

Microsoft Fabric began experiencing widespread issues on Feb 3, 2026 6:53 PM UTC, with the Fabric status page listing DirectLake sync and refresh failures affecting semantic models built on warehouses. Across regions (Americas, Europe, Africa, Asia Pacific, and the Middle East), services were listed as degraded as engineers identified the root cause and warned that a fix was expected by end-of-day Feb 4, 2026 11:59 PM UTC.

Over the following days, the issue expanded to include maintenance-related read-only mode for the Canada Central Fabric cluster during a scheduled window from Feb 7, 2026 2:00 AM UTC to Feb 7, 2026 2:00 PM UTC; users faced limitations on creating or modifying workspaces and potential write errors. Other symptoms affected data-plane operations, warehouse artifacts accessed via SQL endpoints, Notebook artifact queries, and Notebook Display() functionality, across multiple Fabric regions (Americas, Europe, Africa, Asia Pacific, Middle East). The Fabric team issued multiple official updates as the situation evolved. By Feb 26, 2026 4:45 AM UTC, Power BI components across all listed regions had returned to up, and monitoring continued to ensure complete restoration. The incident lasted roughly three weeks with ongoing updates.

Microsoft Fabric experienced a brief partial outage affecting email notification delivery across regions powering Power BI (Middle East, Americas, Europe, Africa, and Asia Pacific). Users may have seen delays in email alerts while backfill queues drained. The incident began at Jan 23, 2026 4:04 AM UTC and was resolved by Jan 23, 2026 9:24 AM UTC, lasting about 5 hours and 20 minutes. Full-service restoration is currently estimated by 11 PM PDT.

The provider published two status updates during the incident, noting that as the dependent service outage was mitigated, backfill queues were being drained and some customers might notice improvements as services returned. By 09:24 UTC, Power BI regional components were back online, with ongoing restoration ETA of 11 PM PDT.

An outage affected Microsoft Fabric's sensitivity label features and related interactions in the Americas Power BI component. Users were unable to apply or manage sensitivity labels and perform interactive actions on labeled reports or artifacts, and email delivery for subscriptions could be delayed. The incident began at Jan 22, 2026 11:40 PM UTC and ended at Jan 23, 2026 3:04 AM UTC, lasting approximately 3 hours and 24 minutes.

The West Europe region of Microsoft Fabric experienced connectivity issues to the SQL Endpoint used for warehouse artifacts, resulting in warehouse data access problems for affected tenants. The outage began on Jan 20, 2026 8:44 AM UTC and persisted for several hours, with service degraded before full restoration. The incident was resolved by Jan 20, 2026 11:19 PM UTC.

The Microsoft Fabric North Europe region experienced a brief outage affecting Notebook operations. The incident began at 4:14 PM UTC and was resolved by 4:24 PM UTC, lasting about 10 minutes. The status was updated twice during the event as engineers investigated and then confirmed restoration of Notebook functionality.

The impact was limited to the North Europe region (Europe / Fabric component); after rapid investigation, the issue was resolved and services returned to normal quickly.

Power BI users with tenants in the South-East Asia-B Primary region experienced issues interacting with Microsoft Fabric services, followed by degraded behavior across Fabric in Asia Pacific. The incident affected two components: Asia Pacific / Power BI and Asia Pacific / Fabric, indicating a regional disruption rather than a global outage.

Microsoft Fabric experienced an issue delivering subscription emails to tenants in the North Central US and East US 2 regions. This degraded the delivery of subscription messages for two components, Americas / Power BI and Americas / Fabric, potentially causing delays or missed emails for affected users. The outage began at Jan 12, 2026 11:59 PM UTC and was resolved by Jan 13, 2026 6:39 AM UTC, lasting about 6 hours and 40 minutes. The incident impacted the two listed regions and components during the window.

Microsoft Fabric customers in the East US region experienced issues with Replicator operations, causing degraded replication for workloads. The issue was first acknowledged at Jan 8, 2026 10:44 PM UTC with engineers beginning investigation in the East US region.

The incident lasted about 3 hours and 5 minutes, with service returning to normal by Jan 9, 2026 1:49 AM UTC. The Fabric team issued two official status updates during the incident, describing ongoing investigation and eventual resolution. The impact was geographically limited to the East US region.

An issue affected Microsoft Fabric users in East US 2 capacities that use Activator rules, with problems triggering actions or viewing live feeds. The outage began at Jan 7, 2026 1:14 AM UTC and was resolved by Jan 7, 2026 4:59 AM UTC, lasting about 3 hours and 45 minutes. The impact was isolated to the Americas Fabric component serving East US 2, with affected users experiencing degraded functionality rather than a complete outage.

Microsoft Fabric's status page listed the incident as a degradation and stated engineering teams were investigating, with an update to follow. A second update at Jan 7, 2026 4:59 AM UTC indicated the issue was resolved and the service returned to normal. The incident involved 2 official status updates during its course.