On July 23, 2026, Azure users around the world began hitting gateway timeouts, DNS failures, and unreachable virtual machines well before Microsoft posted anything on its status page. The first reports reached StatusGator at 15:06 UTC. By 15:28 UTC, StatusGator had sent an Early Warning Signal to subscribers. Microsoft did not acknowledge the incident until 16:29 UTC.
StatusGator’s Early Warning Signal fired 61 minutes before Microsoft acknowledged the outage. Counting from the first user report, the crowd knew 83 minutes before it was official. That is the window you either spend unaware or spend responding.
What happened
Starting around 15:06 UTC, Azure customers reported a wave of failures concentrated in the West US regions but felt in pockets across the globe. The most common symptoms were:
- Azure Front Door 504 gateway timeouts, with the standard “The service behind this page isn’t responding to Azure Front Door” error.
- Microsoft Graph API timeouts, which cascaded into dependent apps and sign-in flows.
- DNS and Application Gateway routing failures, including reports that back-end apps were healthy but unreachable because DNS routing through the app gateway was broken.
- VPN gateway and virtual machine connectivity drops, leaving users unable to RDP into servers or reach internal networks.
- Entra (identity) slowdowns, with admins unable to navigate the Entra admin experience and enterprise apps failing to load.
The problem did not hit every customer equally, but it was widespread enough to affect app services, functions, storage, Redis, PostgreSQL, and identity at once.
Timeline of the incident (UTC)
- 15:06 First outage reports arrive at StatusGator.
- 15:07 to 15:27 Reports accelerate quickly, citing Azure DNS issues, Microsoft Graph API timeouts, and West US connectivity failures across multiple US states and international locations.
- 15:28 StatusGator sends an Early Warning Signal to subscribers, flagging the pattern of unofficial reports before any official acknowledgment.
- 15:28 to 16:28 Report volume continues to climb, spreading from West US into reports from the UK, India, Canada, Brazil, Vietnam, Spain, New Zealand, and more.
- 16:29 Microsoft officially acknowledges the incident on its status page, 61 minutes after StatusGator’s Early Warning Signal.
- 17:51 The last outage reports come into StatusGator, marking the tail end of recovery for affected users.
Impact on users
The reports painted a clear picture of an incident that touched core Azure building blocks rather than a single service. Users described:
- Entire app service footprints going dark, including one report that “All of our app services in US West 2 are down.”
- Serverless and database failures, such as “West US Azure function having connectivity issues with Azure SQL.”
- Identity and API degradation, including “Graph taking excessive amount of time and causing service outages.”
- Network-layer breakage, with one detailed report noting that DNS routing through the application gateway and connections to storage accounts appeared broken while the back-end application itself was healthy.
The downstream effects were real and immediate. One user reported that “ShopGoodwill’s seller portal is down due to an issue with Azure,” a reminder that a single cloud incident ripples out to the businesses and end users built on top of it.
A snapshot of what users reported, from around the world:
- “Azure DNS issue” (Nashville, Tennessee)
- “MS Graph Api is timing out” (Kansas City, Missouri)
- “Azure Front Door service isn’t responding” (Phoenix, Arizona)
- “Can’t connect to Azure VPN gateway” (Ho Chi Minh City, Vietnam)
- “Azure service is down. Any update when it will be up?” (Chandigarh, India)
- “No internet connectivity with Virtual Machines” (Salt Lake City, Utah)
- “Websites hosted on Azure VMs are unresponsive” (Fresno, California)
StatusGator insights
The lead time is the whole story here. StatusGator subscribers were alerted at 15:28 UTC while Microsoft’s status page stayed green until 16:29 UTC. If you were watching only the provider’s page, that was an hour spent unaware your Azure workloads were degraded. Subscribers who got the Early Warning Signal spent that same hour checking dependencies, communicating proactively, and starting mitigation while the provider was still investigating.
Two other patterns stood out in the StatusGator data:
- The reports self-diagnosed the incident. Long before any official root cause, users were pointing at Azure Front Door, Microsoft Graph, DNS, and Application Gateway. Aggregated user reports often surface the shape of an incident faster than a status page can.
- It was widespread but not universal. Reports clustered in West US but appeared from six continents, yet plenty of Azure customers were unaffected. This is exactly the kind of partial, regional incident that official dashboards are slow to reflect, because it does not trip a global “all systems down” threshold.
Lessons learned
Partial and regional cloud outages are the norm, not the exception, and they are the hardest to catch by watching a provider’s status page alone. A few practical takeaways from this incident:
- Do not wait for the official acknowledgment. It lagged reality by more than an hour in this incident. If your monitoring depends solely on the provider’s page, you inherit that delay.
- Watch for symptom clusters, not just outright downtime. The earliest signals here were gateway timeouts, DNS errors, and Graph API slowness, not a clean “service is down” message. Degradation is a leading indicator.
- Map your dependencies. Front Door, Graph, DNS, App Gateway, and Entra are load-bearing for many stacks. Knowing which of your services rely on them turns a vague alert into a targeted response.
- Turn on proactive alerting. Early Warning Signals exist precisely for the window when something is clearly wrong but not yet officially confirmed.
- Keep your own history. Track how often a provider’s acknowledgment lags reality so you can plan around it. You can review Azure’s outage history and the Azure outage map to see status and outage data over time and across regions.
Track Azure status with StatusGator
If you rely on Azure, you should not be the last to know when it is having problems. StatusGator aggregates official status pages and real user reports, then sends Early Warning Signals the moment a pattern emerges, often well before the provider confirms anything.
Ready to stop finding out from your customers? Start monitoring with StatusGator and get early warnings for the services your business depends on.
