[Brian Armstrong: Will Reevaluate Architecture, Future Downtime Should Be Significantly Reduced] Odaily Planet Daily News - Coinbase CEO Brian Armstrong posted on the X platform stating that Coinbase experienced downtime last night due to a failure in multiple cooling units at an AWS data center, which caused overheating in the server rooms. Most systems are designed with redundancy to handle failures in a single AWS availability zone and continue operating normally, but centralized exchanges have not implemented redundancy due to optimizations for latency and client hosting. While it is possible to make the exchange resistant to availability zone failures, this would introduce latency and disrupt client hosting. In light of this incident, these trade-offs will be reevaluated, and at the very least, downtime should be significantly reduced when availability zone migration is required. Thanks to the AWS and Coinbase teams for working overnight to mitigate the issue. A detailed technical summary will be shared later.