Having trouble getting online services to load? You’re not alone. A major outage is currently impacting key parts of the internet, with Microsoft confirming disruptions to Microsoft 365 and Azure while users report issues reaching services from other major providers.
Grammatically speaking, readers of an article on a grammar and spelling checker may be confused because the issue may impact end-user devices and services.
Microsoft service health pages acknowledge problems affecting those experiencing issues with Microsoft 365 and Azure, after some users initially had trouble getting their status monitoring portals to load at all. The symptoms include intermittent sign-in failures, timeouts, and degraded performance with cloud-hosted apps.
At the same time, users are also flagging trouble accessing popular services tied to other cloud platforms, although not all have published their own official incident notices yet. The pattern—simultaneous trouble across multiple companies’ products—hints at the possibility of a broader network or routing incident rather than a single vendor failure. Industry experts often identify issues at backbone transit providers, large-scale Domain Name System hiccups, or Border Gateway Protocol misconfigurations as early warning signs, as these frequently lead to issues across otherwise unrelated platforms.
Services already confirmed to have problems so far include:
- Outlook
- Teams
- OneDrive
- SharePoint
- Azure-hosted applications, particularly where identity services are involved in authentication
- Sluggish dashboards for consumer and developer services
- Stalled API calls
- Streaming interruptions
Not all reports are a 100% outage, but more importantly, the geography- and ISP-spanning scope is a strong indication of a common upstream bottleneck. The stakes are high due to the beam-like nature of cloud usage. According to market analysts at Synergy Research, AWS, Azure, and Google Cloud together account for almost all global cloud infrastructure spend, with Azure hitting up to a quarter share and AWS slightly below one-third in recent quarters, after Microsoft reported signing over 345 million paid seats for Microsoft 365. Even the briefest disruptions will affect workplaces, schools, and public services.
Why multiple internet services can fail at once
Today’s internet is full of dependencies. Regional transit problems, misrouted BGP announcements, or DNS resolver issues can simultaneously incapacitate access to many unrelated brands. Past occurrences follow the same trend: in 2021, a worldwide CDN configuration mistake resulted in a short-term disruption to news sites and government portals, while a 2020 backbone incident disrupted connectivity throughout much of the Midwest U.S.; and after a 2020 routing mistake, a social media giant’s services disappeared from the internet for hours.
Because large clouds interconnect with countless carriers and edge networks, the fault domain can be broader than a single vendor’s data center. That’s why some users can reach a service while others cannot—path selection and caching mean your experience may depend on your ISP, location, and even the sequence of requests your device has cached.
What you can do right now
- Try an alternate connection. If you’re on office Wi‑Fi, test your phone’s cellular hotspot; if mobile data is slow, switch back to your broadband connection.
- Check official status pages and trusted outage trackers for patterns rather than one-off errors.
- If you use corporate accounts, look for IT advisories or temporary workarounds, such as routing traffic through an approved VPN endpoint.
- Avoid drastic account changes. Password resets, app reinstalls, or DNS overhauls rarely fix upstream outages and can create new headaches later.
- If you manage production systems, implement graceful degradation, back off aggressive retries, and extend timeouts to reduce cascading failures.
How providers typically respond during outages
- Activate incident response playbooks quickly.
- Isolate affected regions and reroute traffic.
- Throttle nonessential features to preserve core functionality.
- Provide continuous, transparent updates while engineers validate fixes.
Once stability returns, customers can usually expect a post-incident analysis, including the root cause, the blast radius, and the guardrails added to prevent a repeat.
For now, intermittent access and sporadic performance are probable until the root cause is resolved. Based on previous similar situations, services typically restore in phases, which include, but are not limited to, authentication, core functionality, and ancillary features. Therefore, restoration to partial availability prior to complete normality is normal. We will also keep an eye on network telemetry and provider statements. If your mission-critical cloud applications depend on us, ensure that you have contingency strategies and transparently notify your teams or clients when our services become stable.