Dropbox Moved 600 Petabytes Off AWS! Here's Everything you need to know.
Source: YouTube · Akhil Sharma · published Jun 7, 2026 · 26:15
In 2015, Dropbox made the counterintuitive decision to migrate away from AWS to build their own infrastructure, a move driven purely by massive cost savings at their specific scale rather than cloud failures 0:00.
Key Takeaways:
• While the rest of the tech industry was racing to adopt AWS, Dropbox was the notable exception moving in the opposite direction 0:00.
• By 2017, the company had moved 20% of its user data off Amazon S3, utilizing custom data centers, hardware, and a proprietary software stack to manage hundreds of petabytes of files 0:11.
• The decision was not due to AWS performance issues or an ideological stance against the cloud, but a calculated financial strategy 0:30.
• Running the math revealed that owning their physical infrastructure would save Dropbox tens of millions of dollars annually based on their unique workload and scale 0:36.
Dropbox's strategy highlights a critical inflection point in cloud computing where hyperscale workloads can become cheaper to manage internally than through major cloud providers.
Sources:
Generate CPE Credits
Generate a professional CPE document from this video's transcript.
Estimated credit: 0.5 CPE hours
Estimate uses the video runtime (1 hour ≈ 1 CPE, rounded to the nearest 0.5, minimum 0.5, maximum 2.0). The final amount can be lower after review, never higher.
Topic: Cloud Security. Commonly maps to: Security Architecture and Engineering, Communication and Network Security. Exact CISSP domains are assigned during generation.
Free account. One generation at a time, with a daily limit.
CPEBuddy is independent and not affiliated with or endorsed by ISC2, ISACA, or any certification body. Exports are formatted for common CPE submissions; acceptance is at your certification body's discretion.
Transcript Preview
First 800 characters of the transcript
So, in 2015, every tech company on Earth was racing on to AWS. It would be startups, banks, and even Netflix. But Dropbox was the only company that went in the other direction. They started moving off of it. And by 2017, they had migrated about 20% of their user data away from S3 and onto their own infrastructure. That means their own data centers, their own custom-designed storage hardware, their own software stack for storing what was at the time hundreds of petabytes of user files. And the thing is, they didn't do it because AWS failed them or because of some philosophical objection to the cloud. They did it because they ran the math and figured out that at their specific scale, with their specific workload shape, owning the metal was going to save them tens of millions of dollars a yea…