Cloudbites
Posts
Were you affected by the AWS outage?

Were you affected by the AWS outage?

PLUS: ChatGPT Gets Smarter Memory Management

Lucy Wang
October 25, 2025

In Today’s Cloudbites:

📉 Were you affected by the AWS outage?

🛡️ THE BIG LESSON: Why Backups Need to Be Far Away

⚡ Exciting News: Launching the Zero To Cloud Accelerator!

☁️ Online AI & Cloud Events to Look Forward to

🧠 PLUS: ChatGPT Gets Smarter Memory Management

Read time: 5 minutes

Hi friends, welcome back to Cloudbites

In this newsletter, I’ll share the simple reason why the big AWS outage happened.

But first, I've got something huge for anyone looking to make a career leap: details on the brand new Zero To Cloud Accelerator program are inside!

You'll get to learn the simple cause of the global outage and a major AI update from OpenAI.

CLOUD COMPUTING ☁️

📉 Were you affected by the AWS outage?

The digital world came to a sudden stop on Monday, October 20, 2025.

Amazon’s biggest and most important cloud center, us-east-1 (in Northern Virginia), went offline for over 15 hours.

This event was a harsh reminder that even the world’s largest cloud service can fail, taking a huge part of the internet with it.

The Problem: A Lost Address

The outage was not a hack. It was a technical failure inside AWS, starting with a service called DynamoDB.

Internal System Failure: AWS uses an internal address book (DNS) to find its services. The address for the vital DynamoDB database suddenly went missing inside the network.
The Domino Effect: Since the main database could not be found, all the other necessary tools failed immediately. This included the system that handles user logins (IAM).
Global Spread: Because this one region (us-east-1) controls so many tools globally, the problem quickly spread to affect everyone worldwide.

The Massive Impact: Who Got Knocked Offline?

The outage was huge, with over 17 million user reports of issues around the world.

Popular apps like Snapchat, Roblox, Duolingo, Coinbase, and Venmo were down.
People could not log in, access money, or use financial tools.
Even Amazon’s own devices, like Alexa and Ring security systems, were affected.

🛡️ THE BIG LESSON: Why Backups Need to Be Far Away

After 15 long hours, AWS fixed the address book problem. But the systems came back online very slowly.

Even with the fix, service was delayed because so many requests failed, and it took a long time to process them all. Full service came back slowly.

The Clear Takeaway for Businesses

The main lesson from this outage is simple: You must prepare for the failure of an entire region.

Old Safety Rule Failed: Companies relying on separate rooms (Availability Zones) in the same region were not safe; one core service failure took down all of them at once.
New Safety Rule: Real protection means using a Multi-Region strategy. This requires having a complete, working backup copy of your apps and data in a totally different AWS region.

The goal is to automatically switch to the far-away backup site the moment the main region fails. This is the only way to minimize downtime.

⚡ Exciting News: Launching the Zero To Cloud Accelerator!

I’ve got something special to share with you.

After years of helping thousands of students learn AWS, I’m launching: Zero To Cloud Accelerator.

This 100-day program is designed to take you from zero to job-ready with:

✅ 35 hands-on AWS projects
✅ 100-day roadmap with weekly milestones
✅ Slack community & tech support
✅ Two live career workshops with me
✅ Lifetime access to all Zero To Cloud courses

Plus, if you join our Pro / Elite plan, you'll get:

➕ Live training sessions to learn AWS
➕ Weekly live Q&A sessions
➕ Private “Pro Lounge” discussion room
➕ 1:1 mentorship calls with me
➕ Personalized career roadmap

The accelerator officially starts on January 10th, 2026, but if you enroll now, you’ll get instant access to all course content and projects, so you can get a head start.

💰 Launch Offer: 10% off with code STUDENT10 (limited time)

👉 You can get all the details by clicking here.

[The Elite plan is capped at 10 students. Pro plan is capped at 100.]

P.S. If you’ve already purchased my All-in-One Bundle, you can reply to this email to request an upgrade. You’ll only need to pay the difference

Let's start 2026 off with a bang!

☁️ Online AI & Cloud Events to Look Forward to

#1 AWS Edge Services Immersion Day (November 18, 2025)

This event is a hands-on workshop focused on improving your architecture with Amazon CloudFront and securing your applications at the edge.

Click here to register.

#2 Microsoft Ignite (November 18-21, 2025)

Microsoft's premier conference for developers and IT professionals. The free virtual experience focuses on the future of AI (Copilot), Azure, Microsoft Cloud, security, and hybrid multi-cloud solutions.

Click here to register.

#3 Google Cloud: Grounding your AI Agents (November 13, 2025)

This is a live webinar focusing on how to use Google Maps and Google Search as grounding services in Vertex AI to build more accurate and reliable AI applications.

Click here to register.

ARTIFICIAL INTELLIGENCE 🤖

🧠 PLUS: ChatGPT Gets Smarter Memory Management

OpenAI has quietly introduced a major quality-of-life update for ChatGPT: automated memory management.

💡 Here’s what you should know:

The system will now automatically clear older, less relevant conversation data to prevent context overload and ensure faster performance.
Paid subscribers get the key ability to prioritize and permanently save important memories.
This means you can flag key preferences or instructions you want the AI to always remember, putting you in control of the AI's long-term knowledge about you.

THAT’S A WRAP

Thanks for reading! 😊

P.S. How was today's email? Reply directly with your feedback, or DM me on LinkedIn @LucyWang-