Coinbase has released a post-incident review of its July 14 service outage, saying the disruption was caused by a misconfiguration in a critical network component during a routine configuration update. The company said the incident lasted about 50 minutes and that customer funds remained safe throughout.
According to Coinbase, at 12:34 on July 14 Beijing time, or 12:34 Eastern Time, it deployed a routine configuration change to a shared production Kubernetes cluster that carries core infrastructure services. The update was meant to support a migration to a new service deployment model designed to improve system performance and reliability.
A resource naming conflict slipped past pre-production checks
In the review, Coinbase said a resource naming conflict was not detected by its pre-production safeguards. As a result, the configuration change unexpectedly affected Kubernetes resources tied to the cluster’s Istio Ingress Gateway, making that network component unavailable.
About three minutes later, all inbound traffic to the cluster was interrupted, leaving internal services unable to reach multiple infrastructure components. Coinbase said a large number of asynchronous workflows depend on those infrastructure services, including trade settlement, fund transfers, and bank card authorization, which led to disruption across several business lines.
Retail, institutional and developer services were hit
Coinbase said the affected services included:
- Retail users were unable to complete off-chain trades, deposits, and withdrawals. Some in-progress transactions appeared stuck and were completed after service was restored.
- Coinbase Card debit card transactions temporarily failed. Credit card payments were not affected, but some card management features were unavailable.
- Coinbase DEX suspended on-chain swap services on Base and Solana.
- Institutional clients on Coinbase Exchange and Prime saw transfer and settlement failures or delays.
- Customers of the Coinbase Developer Platform were unable to complete onboarding, fund transfers, and fiat funding services.
Traffic resumed and backlog processing followed
Coinbase said the ingress gateway recovered at 13:20 ET, and the incident was mitigated at 13:23 ET. Systems then began processing the backlog. Most services returned to normal quickly, while the remaining queues were cleared over the following several hours.
Recovery faced two additional obstacles
Coinbase said the recovery process ran into two extra challenges. First, its deployment tooling itself depended on the affected ingress gateway, which meant the normal rollback process could not run and created a circular dependency in which the failure impaired the tools needed to fix it. Second, the company’s break-glass access process worked, but permission checks and manual review added time to the recovery.
Coinbase said the incident did not involve asset security risk. It also said the event exposed areas where large-scale financial infrastructure still needs work, including automated deployment, system dependencies, and disaster recovery mechanisms. The company said it will continue strengthening system resilience to reduce the chance of a similar incident happening again.