If you tried to check your crypto portfolio, make a trade, or use a decentralized app on October 20, 2025, there is a good chance nothing worked. Coinbase had login problems, MetaMask connections failed, and even some blockchain networks slowed to a crawl. The culprit was not a hack, not a blockchain failure, and not a market crash — it was an outage at Amazon Web Services (AWS), the cloud computing giant that powers a surprising amount of the crypto world. Here is what happened, why it matters, and what you can learn from it.
The Basics
Most people think of cryptocurrency as entirely decentralized — that is the whole point, right? No banks, no middlemen, no single point of failure. In theory, that is true. Bitcoin and Ethereum blockchains themselves kept running perfectly on October 20. No blocks were missed, no transactions were lost, and the networks remained secure.
But here is the catch: while the blockchains are decentralized, the services you use to interact with them often are not. The website you visit to check prices, the app you use to trade, the wallet that shows your balance — many of these run on servers owned by Amazon, Google, or Microsoft. When AWS’s US-EAST-1 region experienced a DNS failure starting at around 07:55 UTC on October 20, those services went down too.
Why It Matters
Think of it like a highway system. The roads (the blockchain) are fine, but the on-ramps (the apps and services) are blocked. You cannot get on the highway even though the highway itself is perfectly functional. Research suggests that over one-third of all Ethereum nodes — the computers that verify transactions and maintain the network — are hosted on AWS. When AWS has problems, those nodes go offline, which means fewer on-ramps to the blockchain and slower performance for everyone.
The outage affected Coinbase, Robinhood’s crypto trading, and Base (Coinbase’s Layer-2 network). Decentralized exchanges saw trading volumes drop as price feeds froze. Some users could not access their wallets because the RPC endpoints — the connection points between wallets and blockchains — were running on AWS servers that were down.
Getting Started Guide
So what can you do to protect yourself from the next cloud outage? Here are practical steps every crypto user should take:
Use a hardware wallet. Hardware wallets like Ledger or Trezor store your private keys offline, completely independent of any cloud service. Even if every exchange and every cloud provider goes down simultaneously, your keys remain safe on a physical device in your possession. This is the single most important step for protecting your assets.
Set up multiple RPC endpoints. If you use MetaMask or another browser wallet, you can configure it to connect to different RPC providers instead of relying on a single default. Services like Alchemy, Infura, and public RPC endpoints give you alternatives. If one goes down during an outage, you can switch to another.
Do not keep all your funds on exchanges. The “not your keys, not your crypto” mantra exists for many reasons, and cloud outages are one of them. When Coinbase went down on October 20, users with funds only on the exchange had no way to access or move their assets until the service recovered.
Understand the difference between blockchain and infrastructure. Knowing that a blockchain itself can be perfectly healthy while the services built on top of it struggle is crucial for making informed decisions during crises. Do not panic-sell when you see headlines about crypto outages — the underlying networks are usually fine.
Common Pitfalls
The biggest mistake newcomers make during cloud outages is assuming the blockchain itself is broken. Social media fills with comments about crypto being a failure, but that misunderstands what actually happened. The technology worked as designed; the centralized services built on top of it did not.
Another common pitfall is attempting risky transactions during an outage. When networks are degraded and price feeds are frozen, executing trades or moving large amounts of crypto can result in unfavorable prices or failed transactions. If you can, wait for services to fully recover before transacting.
Finally, avoid falling for scams that emerge during outages. Phishing emails claiming to help you recover access to your account or offering alternative trading platforms spike during these events. Always verify URLs and use only official channels.
Next Steps
The October 20 AWS outage is an opportunity to make your crypto setup more resilient. Start by moving significant holdings to a hardware wallet if you have not already. Configure backup RPC endpoints in your wallet software. Learn how to verify blockchain transactions independently using block explorers like Etherscan, which can function even when some wallet services are down.
The crypto industry is also evolving to address these weaknesses. Decentralized infrastructure networks, often called DePIN, are building alternatives to centralized cloud providers by distributing compute and storage across thousands of independent nodes. As these networks mature, future outages like October 20 should become less impactful — but until then, a little preparation goes a long way.
Disclaimer: This article is for educational purposes only and does not constitute financial advice. Always do your own research before making any financial decisions.
rpc nodes on aws is the real single point of failure everyone ignores
one AWS DNS failure and half of crypto went dark. the blockchain kept producing blocks that nobody could read. peak irony
us_east_1_|_ghost ran a full node specifically because of days like this. everyone laughs at self hosting until coinbase goes blank
blockchain kept producing blocks but you couldnt check your balance because the RPC node was on AWS. decentralization theater exposed in real time
rpc_haunter_ 1/3 of ETH nodes on AWS. the L1 is decentralized but everything you actually interact with runs on 3 cloud providers
blockchains kept producing blocks while the apps on top were down. decentralization working exactly as designed and nobody noticed because the UI was broken
coinbase login issues during an AWS outage is the exact reason self custody exists. one s3 problem and your portfolio vanishes from your screen
the fact that MetaMask connections failed during the AWS outage proves how many dApps secretly run on a single cloud provider. one region goes down and web3 disappears
Selma B. exactly. everyone talks about decentralization until us-east-1 takes a nap and your wallet shows zero balance because the RPC node is hosted on AWS
1/3 of ethereum nodes running on AWS is the real takeaway here. the blockchain stayed up but the on-ramps went down. decentralization is only as strong as its weakest dependency
decentral_fan one third of eth nodes on aws means your wallet balance depends on a beijing to virginia roundtrip. oct 20 proved the chain doesnt matter if the rpc is down
1/3 of ETH nodes on AWS is the dirty secret nobody wants to talk about. blockchain decentralization is theater if the hosting is centralized
one third of ETH nodes on AWS. self custody means nothing when your wallet UI depends on a Virginia data center
Every cycle the infrastructure gets more robust
coinbase metaask and base all dark from one dns failure. blockchains produced blocks normally but nobody could read them. self custody means nothing without your own node
Coinbase MetaMask and Base all dark from one DNS failure. chains produced blocks nobody could read. says everything
the blockchain kept running but you couldnt even check your balance because the RPC node was on AWS. decentralization theater at its finest. everyone preaches self custody until the dashboard goes dark
one aws dns failure and coinbase plus base both went dark even though chains kept running
blockchain kept producing blocks but RPC nodes on AWS meant nobody could read them. decentralization theater
Bear markets are for building — and builders are delivering
The gap between crypto and TradFi is narrowing fast
coinbase, robinhood crypto, and base all going dark from one AWS DNS failure. the gap is narrowing on paper but the infrastructure layer is still deeply centralized
coinbase and base both going dark from one DNS failure proves the user facing layer is where centralization actually hurts