Cloud Outage Analysis

Azure Outage Cost - History, Real Losses & Your Exposure

Updated June 2026 · 8 major incidents documented

Direct Answer

Azure's most recent major outage was 29 October 2025, a roughly 8.5-hour global failure of Azure Front Door (the edge/CDN entry point for many Microsoft services) caused by a faulty data-plane configuration change. It cascaded across Microsoft 365, Outlook, Copilot, the Azure Portal, and Xbox, and took down customer-facing systems at Alaska Airlines, Costco, and Starbucks before engineers rolled the change back. Azure outages tend to be global and authentication- or edge-related rather than single-region, which makes a multi-region or multi-cloud failover plan the main mitigation. Full chronology and SLA credit math below.

Azure Outage History

DateServicesDurationImpactEst. Cost
Oct 2025Front Door (global edge/CDN); cascaded M365, Entra ID, Azure Portal, Xbox~8.5 hoursGlobal; Microsoft 365, Outlook, Copilot, Xbox; Alaska Airlines, Costco, StarbucksNot yet quantified
Jul 2024Azure VMs (CrowdStrike-adjacent)Varies by recoveryWindows Azure VMs with CrowdStrike crashedIncluded in $5.4B CrowdStrike total
Nov 2023Teams, M365, Entra ID~8 hoursMillions of enterprise users globally$50M+ est.
Jul 2023Multiple Azure services~3 hoursDDoS-related degradation$15M+ est.
Jan 2023Storage, VMs in multiple regions~6 hoursAzure Storage degradation$20M+ est.
Sep 2022Azure Active Directory~4 hoursAuthentication failures globally$30M+ est.
Apr 2021Azure DNS~3 hoursDNS resolution failure, cascading to many services$20M+ est.
Sep 2018Multiple regions (lightning)~24 hoursSouth-Central US region outage$10M+ est.

Cost estimates are aggregate customer revenue-at-risk except where noted; Microsoft does not disclose customer impact figures. The October 2025 Front Door outage has no firm published insured-loss estimate, so it is marked "Not yet quantified" rather than guessed; duration and scope are per Microsoft's Azure status updates and ThousandEyes' outage analysis. The CrowdStrike July 2024 figure is Parametrix's $5.4B Fortune 500 aggregate. Updated June 2026.

Azure SLA Credits (Updated 2024)

Monthly Uptime %CreditNotes
99.99%+No credit (SLA met)Standard commitment for most services
99% - 99.99%25% of monthly feeAuto-applied for some incidents since 2024
<99%100% of monthly feeMaximum credit per service per month

Full SLA credit vs actual loss analysis - credits cover only a fraction of business loss.

Frequently Asked

What caused the October 2025 Azure outage?
The 29 October 2025 Azure outage was a roughly 8.5-hour global failure of Azure Front Door, Microsoft's edge and CDN entry point. A faulty configuration change in the Front Door data plane caused nodes across the global fleet to fail to load correctly, cascading to Microsoft 365, Outlook, Copilot, the Azure Portal, and Xbox, and to customer sites including Alaska Airlines, Costco, and Starbucks. It began around 15:45 UTC and was largely resolved by 00:05 UTC on 30 October after Microsoft rolled back the change. No firm insured-loss figure has been published.
How reliable is Azure compared to AWS?
Azure and AWS have comparable uptime records for major incidents. Azure incidents often affect Microsoft 365 and Teams - extremely visible to enterprise users. The CrowdStrike July 2024 outage affected Azure customers running Falcon on Windows Azure VMs, creating a compound incident included in the $5.4B Fortune 500 loss estimate.
How do Azure SLA credits work?
Azure credits are 25% of monthly service fees when uptime falls to 99%-99.99%, and 100% below 99%. In 2024, Azure updated to automatically apply some credits for verified incidents. Credits are Azure billing credits, not cash refunds, and are capped at the monthly fee for the affected service.

Related

AWS Outage History

Compare AWS reliability

SLA Credit Math

Credit vs actual loss asymmetry

All Major Outages

Full incident database

Calculate Your Exposure

Your Azure outage cost

Updated 2026-04-27