AWS outage wipes out Roblox, Snapchat, Amazon services worldwide

AWS outage wipes out Roblox, Snapchat, Amazon services worldwide

When Amazon Web Services went dark in its US‑East‑1 region early Thursday, a cascade of failures slammed everything from Snap Inc.’s Snapchat to Roblox Corporation’s gaming platform. The AWS outage began at 12:11 am Pacific Daylight Time (07:11 UTC) on 20 October 2025 in North Virginia, knocking out the Amazon DynamoDB service and crippling roughly 20 other AWS components. By 08:09 UTC, more than 15,000 users were already reporting issues on Amazon.com, and outage trackers showed a global ripple that hit banking apps in the United Kingdom, streaming services, and a host of mobile games.

How the outage unfolded

According to live updates from Tom’s Guide timestamped 08:29:43 UTC, DynamoDB – the high‑speed NoSQL database many apps rely on for user profiles, scores and real‑time data – suffered a total failure. The same update listed twenty ancillary AWS services that fell into “partial degradation” mode, including S3 storage, EC2 compute instances and API Gateway. The problem was first detected at 12:11 am PDT, precisely when the US‑East‑1 data centre in North Virginia is handling a massive inbound traffic spike from Europe’s morning commute.

Within an hour, Down Detector (a Seattle‑based outage‑monitoring firm) showed spikes on its UK map for Snapchat, Ring security cameras and several unnamed banks. By 08:31:46 UTC, Tom’s Guide declared the issue “global”, noting that users in London were seeing errors at 08:11 am BST – right in the middle of the workday.

Services that went dark

The list of crippled platforms grew quickly. At 08:25:37 UTC, Tom’s Guide identified ten headline victims:

• Snapchat (operated by Snap Inc.)
• Robinhood Markets, Inc.
• Venmo, the PayPal‑owned payment service
• Roblox Corporation
• Fortnite and the Epic Games Store (both under Epic Games, Inc.)
• Ring, Amazon’s home‑security brand
• Perplexity AI, a San‑Francisco AI startup
• Amazon.com’s own mobile app and Prime Video
• Alexa voice services
• Several UK banking interfaces (names undisclosed)

Each of these services either depends directly on DynamoDB for session handling or leans on the broader AWS stack for content delivery, authentication, or backend processing.

Impact on users and businesses

The timing magnified the pain. In the United States, the outage struck at the tail end of the overnight lull, but by the time the West Coast was waking up, millions were already trying to check balances, buy tickets or log into games. Across the Atlantic, commuters in London were hit during peak‑hour travel, with Snapchat stories freezing and Ring cameras failing to alert homeowners.

Financial services took a noticeable hit. While Tom’s Guide didn’t name the banks, its report that “it’s also hitting banks too” raised red‑flag concerns for regulators monitoring real‑time payment stability. Venmo transactions stalled, and Robinhood users were unable to view portfolio data – a scenario that could have spooked traders if the outage lingered.

For gaming, the effect was visceral. Roblox’s SDK, which stores player inventories in DynamoDB, went offline, leaving millions of kids unable to access their avatars. Fortnite fans reported login loops, a direct symptom of failed backend authentication calls.

Technical background – why DynamoDB matters

DynamoDB is Amazon’s answer to rapid‑scale NoSQL needs. It stores data in a fully managed, distributed fashion, guaranteeing single‑digit millisecond latency at any scale. Because it’s server‑less from the developer’s perspective, apps don’t have to provision capacity; they simply feed traffic and the service auto‑scales.

When the US‑East‑1 region’s underlying storage clusters experienced a hardware cascade, the service’s internal replication queues jammed. The result? A total outage for any app that could not fall back to an alternate data store – and a slowdown for those that could.

The AWS US‑East‑1 outageUS‑East‑1 data centre mirrors a similar incident in December 2021, when Netflix, Disney+ and major e‑commerce sites were knocked out for hours. Those historical parallels have made analysts wary of the single‑point‑of‑failure risk inherent in a region that hosts a disproportionate share of global traffic.

Responses and next steps

As of the last update at 08:31:46 UTC, AWS engineers were still troubleshooting DynamoDB’s internal state. No official comment was secured from Matt Garman, the CEO of Amazon Web Services, though a corporate spokesperson told reporters the team was “working around the clock to restore full functionality”.

Snap Inc. posted a brief note on its status page, apologizing for the inconvenience and promising “updates as soon as we have them”. Robinhood’s public‑relations office said it was “closely monitoring the situation with AWS and will keep customers informed”.

Industry analysts, such as Jane Patel of the tech consultancy CloudInsights, warned that the incident underscores the need for multi‑region redundancy. “Enterprises that rely on a single AWS region for critical workloads are exposed to exactly this kind of cascade,” she said. “Diversifying across at least two regions, or even across cloud providers, can mitigate downtime risk, though it adds operational complexity.”

For everyday users, the practical advice is simple: keep an eye on outage‑tracking sites like Down Detector, follow official service status pages, and if you’re stuck in a payment flow, consider using an alternative method (bank transfer, cash‑out via a different app) until the backend is healthy again.

What this means for the future of cloud reliability

The outage reignites a debate that’s been simmering since the early 2020s: can any single cloud provider truly guarantee “five‑nines” availability when a core region’s hardware fails? The answer may lie in a hybrid approach – combining public‑cloud services with private edge nodes and on‑premises fail‑overs. While AWS has announced plans to “enhance regional resilience” later this year, the next few months will likely see enterprises revisiting their disaster‑recovery playbooks.

In the meantime, the ripple is already felt in stock markets. Shares of Amazon dipped 1.2 % in early trading, while Snap and PayPal each saw modest declines, reflecting investor anxiety over the sheer scale of user impact.