Use Cases
Software Products MSPs Schools Development & Marketing DevOps Agencies Help Desk
 
Internet Status Blog Pricing Log In Try IsDown for free now

Outage in Fivetran

Multiple Twitter Ads and Twitter Organic Connections Failing with error "HTTP 503 Service Unavailable"

Resolved Minor
March 30, 2026 - Started 1 day ago - Lasted about 24 hours
Official incident page

Incident Report

Summary AI Generated

Multiple Twitter Organic and Twitter Ads connections in Fivetran failed for 23.5 hours due to "HTTP 503 Service Unavailable" errors caused by the X API not returning rate limit information in response headers, which led to repeated 429 rate limit errors and sync failures. The issue was resolved when X fixed their API problem, and Fivetran deployed a hotfix to automatically reschedule connections that encounter account-level rate limits without proper response headers to prevent future sync failures.

The issue has been identified and we are working to resolve it.

Need to monitor Fivetran outages?

  • Monitor all your external dependencies in one place
  • Get instant alerts when outages are detected
  • Be the first to know if service is down
  • Show real-time status on private or public status page
  • Keep your team informed
Latest Updates ( sorted recent to last )
RESOLVED about 15 hours ago - at 03/31/2026 05:25PM

This incident has been resolved. We have observed that error rates have returned to normal levels, and the Twitter Organic and Twitter Ads services are operating as expected.

Incident Summary

Description:
We identified an issue affecting multiple Twitter Organic and Twitter Ads connections, which were failing with the error: “HTTP 503 Service Unavailable.”

Timeline:
The issue began on March 30th at 00:00 UTC and was resolved on March 31st at 16:30 UTC.

Cause:
The X API stopped returning rate limit information in the response headers. This caused Twitter Organic and Twitter Ads connections to encounter 429 errors, leading to sync failures.

Resolution:
X has resolved their API issue. Additionally, we deployed a hotfix to ensure that if the source hits account-level rate limits and does not return response headers, it will no longer cause sync failures. Instead, the connections will be automatically rescheduled.

MONITORING about 16 hours ago - at 03/31/2026 03:54PM

The issue appears to have been resolved from the source side. Rate limit errors have subsided, and affected connectors have returned to normal sync operations

IDENTIFIED about 20 hours ago - at 03/31/2026 11:56AM

We are continuing to monitor the status page, as there are no latest updates from X.
In the meantime, we are exploring potential workarounds to help mitigate the issue until it is fully resolved on the source side.

IDENTIFIED 1 day ago - at 03/31/2026 03:56AM

We are still seeing the issue at the source end and are continuing to monitor their status page. We have reached out to X support, and X has acknowledged degraded performance on their status page: https://docs.x.com/status.

IDENTIFIED 1 day ago - at 03/30/2026 06:09PM

The issue appears to be related to missing rate limit headers in the response. As a result, requests are repeatedly hitting HTTP 429 (rate limit) and, in some cases, resulting in HTTP 503 errors. We have reached out to X support, and X has also acknowledged degraded performance on their status page: https://docs.x.com/status

IDENTIFIED 1 day ago - at 03/30/2026 05:55PM

The issue has been identified and we are working to resolve it.

The Status Page Aggregator with Early Outage Detection

With IsDown, you can monitor all your critical services' official status pages from one centralized dashboard and receive instant alerts the moment an outage is detected. Say goodbye to constantly checking multiple sites for updates and stay ahead of outages with IsDown.

Start free trial

No credit card required · Cancel anytime · 6020 services available

Integrations with Slack Microsoft Teams Google Chat Datadog PagerDuty Zapier Discord Webhook