Mercado Livre faces global instability this Monday, impacting user searches and login

Users of Mercado Livre faced service disruptions starting at 8:40 AM on Monday, October 20, 2025, across several regions in Brazil. The instability affected access to the website and app, preventing product searches and order viewing. The issue stems from a failure in the infrastructure of Amazon Web Services (AWS), the cloud provider used by the platform.

Reports on the monitoring site Downdetector show complaint spikes above normal, with 53% related to the mobile app, 35% to the website, and 13% to login issues. AWS confirmed an operational disruption in the US-East-1 region in the United States, impacting global services dependent on its data centers.

The platform, one of the largest in e-commerce in Latin America, records millions of daily accesses, and the incident occurred during peak morning hours. AWS technical teams are working to fully restore stability.

Causes of the infrastructure disruption

The failure began around 3:30 AM Brasília time at the AWS data center in Northern Virginia. This location hosts critical controls for services like storage and content distribution, causing a cascade effect on connected platforms.

In Brazil, the impact focused on essential Mercado Livre features, such as navigation and transactions. AWS reported increased error rates in multiple services, including databases and networks.

Users in other countries, such as the UK and the US, also reported similar issues in dependent apps. The company activated parallel recovery pathways to mitigate the effects.

User reports and affected features

Sellers and buyers shared experiences of extreme slowdowns and loading errors. A seller from São Paulo mentioned an inability to access recorded sales, halting daily operations.

The app allowed initial access but failed to return search results. Payments via Mercado Pago faced intermittent rejections, complicating transactions.

• 53% of reports involve the app, with complaints of crashes.
• 35% cite the website, focusing on pages that fail to load.
• 13% refer to logins, with persistent error messages.

These Downdetector data highlight a volume above the usual hourly average.

AWS-dependent services impacted

Platforms like Canva and iFood reported similar complaints, with issues in image editing and delivery orders. Amazon’s Prime Video experienced slowdowns in streaming, affecting morning viewership.

Roblox and Fortnite faced interruptions in online gaming, halting active sessions. The Alexa virtual assistant stopped processing voice commands on connected devices.

These services share dependency on AWS’s US-East-1 region for hosting. Partial recovery occurred around 7 AM, but request backlogs persist in some cases.

History of cloud platform outages

Similar disruptions occurred in September 2024, when Mercado Livre faced instabilities for over two hours, accumulating thousands of reports. That incident focused on financial transactions, resolved after server updates.

In 2023, a global AWS outage affected e-commerce platforms in Latin America, lasting about 90 minutes. The incident led to improvements in network redundancies, but vulnerabilities remain during demand peaks.

In the current episode, the estimated duration ranges from 30 to 60 minutes for full restoration. AWS is monitoring traffic to prevent immediate recurrences.

Recovery measures implemented

AWS teams deployed traffic redirects to alternative data centers. This accelerated the normalization of 70% of affected services by 9 AM.

Mercado Livre advised users to try accessing via an alternative browser or await updates. Local caching tools aided partial access during the outage.

The platform notified partners about the incident, prioritizing high-volume sellers. Internal tests confirm stability in secondary regions.

Future stability outlook

Firmware updates in data centers aim to reduce risks of cascading failures. AWS plans to expand capacity in South America for greater local resilience.

Users can monitor status via AWS’s official dashboards. Incidents like this underscore the importance of multi-cloud backups for businesses.

Full normalization is occurring gradually, with a focus on preventing disruptions during business hours.