Fivetran Outage History

Fivetran's Twitter Ads and Twitter Organic connectors experienced sync failures for 17 hours due to the Twitter API returning HTTP 429 and 503 errors. The issue affected data synchronization from Twitter services, causing connectors to fail when attempting to retrieve data. Fivetran resolved the incident by deploying a hotfix that better handles API errors from Twitter and skips endpoints that continue to fail after retries.

Fivetran's transformation jobs (dbt core and Quickstart) were failing with "UnsatisfiedDependencyException" errors due to a recent code change. The incident lasted 1.5 hours from April 10, 2026 at 00:40 UTC to 01:30 UTC, affecting the Transformations component. Engineering resolved the issue by identifying the problematic code change and reverting it, restoring normal transformation operations.

Fivetran experienced a bug that caused Amazon Selling Partner connections to fail when syncing the all_listings_report endpoint, with connectors throwing a null context error. The issue affected Amazon Selling Partner integrations for 3.1 hours on April 8, 2026. Fivetran implemented a fix to resolve the null context bug, and affected connectors returned to normal syncing operations.

Fivetran's Twitter Ads and Twitter Organic connections experienced sync failures for 6.9 hours due to HTTP 503 (Service Unavailable) errors from Twitter's API endpoints. The issue was caused by Twitter's '/stats/jobs/accounts/' endpoint returning service unavailable errors, preventing data synchronization. Fivetran resolved the incident by deploying a hotfix that handles 503 errors and implements retry logic for failed endpoints.

Fivetran's Awin connections experienced sync failures due to 500 internal server errors from the Transaction endpoint, caused by an issue on Awin's source side. The incident lasted 5.1 hours and affected all Awin connector syncs. Fivetran resolved the issue by implementing a workaround that skips the problematic Transaction endpoint, allowing connections to sync successfully while coordinating with Awin support for a permanent fix.

A recent deployment caused SQL Server (Teleport) and Epic CSA connectors to fail with a "SYSNAME" data type error when processing VIEW tables. The issue affected data syncs for these connectors for 3.6 hours until the problematic deployment was rolled back and a hotfix was applied to restore normal operations.

Multiple Twitter Ads connections in Fivetran experienced failures due to HTTP 500 Internal Server Errors from the 'campaigns' endpoint. The issue affected Twitter Ads data synchronization for 7.4 hours. Fivetran implemented a fix and workaround to resolve the endpoint failures, restoring normal connection functionality.

Reddit Ads connections in Fivetran experienced sync failures due to HTTP 500 Internal Server errors from Reddit's API, specifically affecting the report creation endpoint across multiple connectors. The issue lasted 4 hours on April 1st and was resolved when Reddit fixed the API errors on their end. Fivetran also implemented a hotfix to prevent similar failures from causing sync disruptions in the future.

Fivetran experienced delays in log delivery to customers' external logging systems due to an outage in the Splunk Log API, which caused a buildup of messages in Fivetran's internal queue and impacted delivery across other logging services. The team deployed a fix to prevent delays to other logging services while Splunk resolved their API issue. The incident lasted 41 minutes and was fully resolved on April 1, 2026, at 02:19 UTC.

Fivetran experienced an issue with Pipedrive connections where syncs were failing due to a null pointer exception in the date handling code. The problem was caused by a configuration issue that affected existing connections and lasted approximately 5 hours and 14 minutes. Fivetran applied a fix to improve error handling and prevent the issue from recurring.