In the digital world, almost everything you use—from streaming Netflix to checking TikTok—runs on giant, unseen computer networks called the "cloud." The biggest of these networks is Amazon Web Services (AWS). When AWS goes down, it’s like the internet’s power grid failing.
The recent, massive AWS outage was a major wake-up call, shutting down services for countless companies, including the parts of the HubSpot platform.
HubSpot is a key tool for businesses, especially the marketing and sales teams, to manage their customer data, websites, and emails. Because HubSpot runs its essential systems on AWS, when Amazon's network broke, a portion of HubSpot's services followed.
The outage, which happened on Monday, October 20, 2025, caused major headaches:
The main issue started in a crucial AWS data center region known as US-EAST-1, located in North Virginia, USA.
Because so many global systems, including core parts of AWS itself, rely on this single region to work, the HubSpot outage wasn't just a US problem; it created a global ripple effect:
Essentially, when the main engine room (US-EAST-1) failed, the whole global ship slowed down. Even customers far away experienced slow speeds and broken connections as their systems struggled to communicate with the core services that were stuck.
Amazon traced the original issue back to a failure in their Domain Name System (DNS) resolution within the US-EAST-1 Region. Think of DNS as the internet's phonebook: it translates human-readable website names (like "HubSpot.com") into the numeric addresses (IP addresses) that computers need to find them. A problem in the system that monitors Amazon’s network traffic load balancers triggered the DNS failure. This meant huge parts of the AWS network lost their ability to correctly route traffic and talk to each other, leading to chaos across various cloud services inlcuding Snapchat, Netflix / Disney+ / Prime Video, Signal messaging app, Canva and some UK banks. Even the Xero accounting platform was effected by outages and sluggishness.
After many hours of work, applying fixes, and slowing down new requests to reduce the load, the AWS engineering teams managed to get the critical systems back online.
As of now, the problem is close to fully resolved. HubSpot confirmed that all their systems are operational and working normally, and importantly, they did not lose any customer data during the incident.
HubSpot has notified that there are still some outages with reporting, dashboards and HubSpot apps in North America. A full list of HubSpot's outage status can be seen online.
This recurring nightmare shows that companies running on the cloud need a better disaster plan. For platforms like HubSpot, here are the key strategies to stay up when AWS goes down:
Don't Use Just One Region: Most companies use multiple Availability Zones (AZs) within one region for backup. If one region melts down, they can instantly switch all traffic to the other.
Be Independent of the 'Manager': The systems that manage and set up AWS resources—the "control plane"—often fail first. Companies should design their apps to be able to run for a few hours even if they can't talk to the main AWS management tools, relying on locally cached data (data saved right there).
Test Your Disaster Plan: It's not enough to have a backup system; you have to know it works. Companies must run "chaos drills"—intentionally breaking their own system or simulating an outage—to make sure their emergency switchovers actually work in real time.
Use Two Clouds (If You Can Afford It): For the absolute most critical services, some large companies are moving to a multi-cloud strategy, running parts of their infrastructure on both AWS and Google Cloud Platform (GCP) or Microsoft Azure. This is expensive and complex, but it guarantees that a failure at one giant provider won't shut you down completely.
For now, HubSpot appears to be intermittenly running and no doubt there are some infrastructure conversations going on right now to discuss mitigation in the future.
It's handy to remember that Salesforce has also had its fair share of outages in the past couple of years - most recently in June of 2025 when it was down for almost a whole day.