Shared infrastructure · who goes down together

Cloud outage cascade tracker

Most of the web runs on a handful of providers. When one has an outage, dozens of sites can go down at once — even when your own code is fine. AWS alone is shared by 28 of the major sites we track. This page maps 97 dependencies across 8 providers and flags any that are reporting problems right now.

⚠ Cloudflare is reporting an outage now. 16 major sites depend on it — also reporting problems: Cloudflare.

AWS

Cloud host28 major sitesNo incident

Cloudflare

CDN16 major sitesOutage now

Fastly

CDN10 major sitesNo incident
GitHubNetflixRedditSpotifyStripeTwitchTwitter / Xnewrelic.comsteampowered.comtwitter.com

Microsoft Azure

Cloud host10 major sitesNo incident
GitHubLinkedInMicrosoft TeamsMinecraftXbox Livechatgpt.commicrosoft.comoffice.comopenai.comskype.com

Google Cloud

Cloud host9 major sitesNo incident

Google Sign-In

Login9 major sitesNo incident

Stripe

Payments9 major sitesNo incident

Akamai

CDN6 major sitesNo incident

Dependency relationships are based on publicly documented infrastructure and provider usage; providers change and most sites use several, so this is a guide to likely shared-infrastructure impact, not a real-time per-site dependency audit. Live status is shown only for incidents in our feed — a provider with no incident reads “No incident”, never a guess. For the authoritative answer, check each provider's own status page.

How a cloud cascade happens

When you load a website, you're rarely talking to just that company's servers. The page is served through a CDN (Cloudflare, Fastly, Akamai), the app runs on a cloud host (AWS, Azure, Google Cloud), logins may route through a third party, and payments often go through Stripe. Each of those is a shared dependency — used by thousands of other sites at the same time.

So when AWS's us-east-1 region has a bad hour, it doesn't take down one site — it takes down a slice of the internet. Streaming, gaming, banking apps and dashboards all fail together, and each company's on-call engineers spend the first ten minutes assuming the problem is their own code. The cascade is the tell: when many unrelated sites break at the exact same moment, the cause is almost always upstream.

Frequently asked questions

What is a cloud outage cascade?
Most of the web runs on a small number of shared providers — AWS, Cloudflare, Google Cloud, Azure and a few CDNs. When one of them has an outage, every site that depends on it can go down at the same time, even though each site's own code is fine. That simultaneous, provider-driven failure is the cascade.
How do I know if AWS (or another provider) is down right now?
This page overlays our live incidents feed onto the dependency graph: if a provider is reporting an outage, it's flagged "Outage now" and the sites also reporting problems are highlighted. For the authoritative source, always check the provider's own status page too — we link each provider to its live WebsiteDown status page.
How do you know which sites run on which provider?
The relationships here are based on publicly documented infrastructure and provider usage (engineering blogs, case studies, DNS/CDN headers). Providers change over time and a single site often uses several, so treat this as a guide to likely shared-infrastructure impact — not a real-time, per-site dependency audit.
My site is down — is it my code or my cloud provider?
Check whether the provider you depend on is flagged here, and whether other sites on the same provider are reporting problems at the same time. If they are, the outage is likely upstream — not your code. If you monitor your site with WebsiteDown, you can declare its providers and get told when an upstream provider outage is the likely cause.

Know when it's the cloud, not your code

Monitor a site with WebsiteDown, declare the providers it depends on, and get an email when an upstream provider outage — not your deploy — is the likely cause.

See monitoring plans →