AWS is having issues today because of ongoing network connectivity problems in one of its core regions, which are disrupting several internal services and causing errors or slowdowns for many apps that rely on AWS. It is a partial outage focused on specific services and regions rather than a total global shutdown.

Quick Scoop

  • What’s happening right now
    • There has been a spike in outage reports for AWS in the last day, with many users seeing timeouts, login failures, or “unable to connect” errors in apps that run on AWS.
* AWS’s own status/health pages list active events describing **network connectivity issues** that are affecting services like DynamoDB, SQS, and others in at least one region (notably us‑east‑1).
  • Why it looks bigger than “just AWS”
    • When AWS has problems in a major region, it hits a lot of big-name platforms at once (social media, APIs, internal business tools), so it feels like “half the internet is down.”
* Outage trackers like Downdetector show parallel spikes for X, Cloudflare, Grok, and AWS, which makes it look like everything is broken, even if each provider is having a slightly different issue.

What’s likely causing the outage

From current and recent incidents, the most plausible causes are:

  1. Network/connectivity issues in a key region
    • AWS has acknowledged ongoing network connectivity problems impacting multiple services in at least one region, which usually means routers, links, or internal networking equipment are degraded or misbehaving.
 * When this happens in us‑east‑1 (Northern Virginia), it often has outsized impact because so many companies centralize workloads there.
  1. Dependency on internal services (DynamoDB, DNS, queues)
    • Many AWS-based apps depend on shared building blocks like DynamoDB (databases), SQS (queues), and DNS resolution; if those slow down or fail, apps time out even if the app’s own servers are technically “up.”
 * A similar large AWS outage in 2025 was traced to an issue involving DynamoDB and DNS in us‑east‑1, which cascaded through thousands of services worldwide.
  1. “Blast radius” and architecture choices
    • A lot of companies still concentrate their workloads in a single AWS region or rely heavily on us‑east‑1, so any fault there can ripple across banking apps, SaaS tools, games, and more.
 * Even companies that claim to be multi‑region or multi‑cloud can still be tied to one region for databases, identity, or control planes, so their failover does not always work smoothly during an AWS incident.

How this shows up for you

Users and admins are typically seeing:

  • Timeouts, 5xx errors, or “Unable to reach server” messages when loading websites or apps that run on AWS.
  • Background tasks (payments, notifications, file uploads, AI calls) stuck in “pending” or silently failing due to queue and database issues.
  • Status pages from affected services pointing to “issues with our cloud provider” or specifically referencing AWS or a single region.

What you can do right now

For regular users:

  • Refresh sparingly; hammering reload can create duplicate actions or failed transactions when backends are unstable.
  • If money or critical actions are involved (banking, orders), wait for official “all clear” from the service and then double‑check recent transactions.

For engineers/admins:

  1. Check official status and scope
    • Look at the AWS global health/status page and your chosen region’s incident notes to confirm which services and regions are affected.
 * Cross‑check with outage trackers (like Downdetector) to see whether your users’ region lines up with current spikes.
  1. Mitigate where possible
    • If your architecture supports it, route traffic to a healthy region and spin up capacity away from the incident region.
 * Temporarily degrade non‑critical features that rely on affected services (e.g., async features that use SQS or secondary databases) so core flows stay responsive.
  1. Communicate with users
    • Post a clear banner or status update explaining that an upstream cloud incident is affecting availability and that data integrity is being protected as the priority.
 * Provide realistic expectations: AWS often restores core functionality within hours, but cleanup of queues, backlogs, and stuck jobs may take longer.

Broader context and lessons

  • Major AWS outages like the 2025 event showed how a fault in one region plus a core service (like DynamoDB and DNS) can “snap the internet’s backbone” for a few hours.
  • Experts consistently recommend:
    • Designing for regional failure (multi‑region or multi‑cloud, tested regularly).
    • Minimizing single points of failure in shared services like databases, identity, and DNS.
    • Investing in clear outage communication and independent monitoring, not solely relying on the cloud provider’s view.

Bottom line: AWS is “down” today in the sense that key services in at least one region are suffering network issues, which is cascading into visible outages across many apps, but it is a partial, region‑scoped incident rather than a permanent or total failure.

Information gathered from public forums or data available on the internet and portrayed here.