Fivetran Outage History

This incident has been resolved. We have observed that instance rates are returning to normal levels and affected connections are syncing successfully. Incident Summary Description: We identified an issue with the HubSpot connection, which resulted in syncs failing with "HTTP 500 Internal Server Error" Timeline: This issue began on 8th May 2026 at 12:07 AM UTC and was resolved on 8th May 2026 at 01:32 AM UTC. Cause: The issue was caused by the third-party, event-definitions endpoint failing with "HTTP 500 Internal Server Error". HubSpot has also updated their status page, confirming the regional degradation due to the AWS outage. Resolution: The issue was automatically resolved from the third-party side without any intervention from the Fivetran end.

This incident has been resolved. We have observed that instance rates are returning to normal levels and affected connectors are syncing successfully. Incident Summary Description: We identified an issue for Klaviyo, which resulted in syncs failing with reconnect tasks thrown on the dashboard. Timeline: This issue began on 7th May 2026 at 8:00 AM UTC and was resolved on 7th May 2026 at 11:50 AM UTC. Cause: Received HTTP 401 failure for the Template endpoint. Resolution: We have added a skip logic for the Template endpoint that disables/does not allow syncing the email_template table when the endpoint is encountering the 401 issue, but allows all other tables to sync normally instead of failing the sync.

We have resolved this incident.

Description: We identified connection issues affecting dependency retrieval during the build process for Hybrid Connector images. Specifically, requests to external package sources were failing, resulting in build disruptions. Impact: Customers using Hybrid Deployment may have experienced failures when attempting to build or update connector images during this period. Timeline: Start Time: 2026-05-04 13:43 UTC End Time: 2026-05-04 14:57 UTC Root Cause: The issue was caused by an inability to fetch required dependencies from Ubuntu Launchpad and the Ubuntu archive, which prevented successful image builds. Resolution: We mitigated the issue by reverting to stable, previously built images, restoring normal functionality for Hybrid Connector deployments.

This incident has been resolved. Incident Summary Description: We identified an issue where the connector SDK connection state is not advancing, resulting in the same data sync for the last few days. Timeline: This issue began on April 29, 2026, at 02:08 UTC and was resolved on May 4, 2026, at 13:00 UTC Cause: A recent code change caused the issue. Resolution: Reverted the changes to resolve the issue.

Instance rates have returned to normal levels, and affected connections are now syncing successfully. Incident Summary Description: We identified an issue impacting HubSpot connections, where syncs started failing with the error "Your HubSpot source is not responding at this time. We'll retry the sync later." Timeline: The issue began on May 1, 2026 at 00:10 UTC and was resolved on May 1, 2026 at 03:30 UTC Cause: The failures were caused by intermittent 502/504 errors returned by the HubSpot events API endpoint, which led to sync disruptions. Resolution: The issue resolved automatically without any intervention from the Fivetran side.

Incident Summary Description: We identified an issue where services were intermittently failing with the error: "UNAUTHENTICATED: Token", resulting in connector sync failures during specific time windows. Timeline: This issue began on 2026-04-30 at 10:00 UTC and was resolved on 2026-04-30 at 16:30 UTC. Customers might have experienced failures around scheduled deployment times (10:00, 13:00, and 16:00 UTC), with recovery shortly after. Root Cause: A recent internal change led to intermittent timeouts during authentication token validation, which triggered "UNAUTHENTICATED: Token" errors and caused some syncs to fail. Resolution: We implemented a fix to ensure sufficient time for token validation, preventing further sync failures. All services are now operating normally.

Instance rates have returned to normal levels, and affected connectors are now syncing successfully. Incident Summary Description: We identified an issue with Qualtrics connections where syncs were failing with a "Ticket export failed" error. Timeline: The issue began on April 28, 2026 at 11:30 AM UTC and was resolved at 4:10 PM UTC. Cause: The failures were caused by an issue with the Ticket Export endpoint on the Qualtrics side, leading to sync disruptions. More details: https://status.qualtrics.com/incidents/wj65y7rwfx2j Resolution: The issue was resolved once Qualtrics fixed the problem on their end.

This incident has been resolved. We have observed that instance rates are returning to normal levels and affected connectors are syncing successfully. Incident Summary Description: We identified an issue for SharePoint and Microsoft Lists connections which resulted in syncs failing with a "Authentication failed: The client secret or certificate has expired" error. Timeline: This issue began on 2026-04-27 at 22:20:00 UTC and was resolved on 2026-04-28 at 09:30:00 UTC Cause: The issue occurred due to Microsoft OAuth app credentials expiration.Due to this internal workflows failed to rotate the secret keys. Resolution: A fix has been implemented to update these secrets.

This incident has been resolved. We have observed that instance rates are returning to normal levels and affected connectors are syncing successfully. Incident Summary Description: We identified an issue for Amazon Selling Partner connections which resulted in syncs re-scheduling due to the rate limits on the GET_SALES_AND_TRAFFIC_REPORT. Timeline: This issue began on 2026-04-24 at 09:00:00 UTC and was resolved on 2026-04-27 at 14:15:00 UTC. Cause: We are experiencing the rate limits issue from the third party causing our connectors to reschedule for the GET_SALES_AND_TRAFFIC_REPORT. Resolution: A fix has been implemented to skip the tables associated with this report and allowing the connections for a successful sync.

Incident Summary: Description: QuickBooks connections were failing due to the source-side maintenance window. Timeline: April 25, 2026, from 9:00 PM to 11:30 PM PT Cause: Source side maintenance Resolution: Automatically resolved after the maintenance window is over.

Incident Summary: Description: Multiple Twitter Ads connectors failed due to the Twitter Ads API returning persistent server errors (HTTP 500) on reporting endpoints. Timeline (UTC): Started at 25.04.2026 21:00 and Resolved 26.04.2026 11:50 Cause: The Twitter Ads API experienced an outage, returning repeated HTTP 500 Internal Server errors on the stats reporting endpoint (POST URI https://ads-api.twitter.com/12/stats/jobs/accounts/). Resolution: Deployed a fix/workaround that enables the connector to gracefully handle HTTP 500 errors from the Twitter Ads API. Affected accounts are now skipped with a customer-facing warning on the connector dashboard, sync progress is preserved for successful accounts, and skipped data is automatically retried on the next sync.

Syncs have returned to normal and Azure looks to have completed the rollout of their fix. Azure will have a post mortem posted to their status page for the root cause of this incident: https://azure.status.microsoft/en-us/status

We have observed that instance rates have returned to normal levels, and the affected connectors are now syncing successfully. Incident Summary Description: We identified an issue with Recharge connections, where syncs were failing with a 502 error. Timeline: The issue began on 2026-04-23 at 15:45 UTC and was resolved at 17:38 UTC. Cause: The failures were caused by the intermittent unavailability of the Recharge API, leading to sync disruptions. More details are available on the recharge's status page: https://status.getrecharge.com/incidents/cbz3mcgf2lx5 Resolution: The issue was resolved automatically once the Recharge API resumed normal operation

Description: We identified an issue affecting ServiceTitan connectors, which resulted in syncs failing with an HTTP 500 error. Timeline: This issue began on 2026-04-23 at 17:55 UTC and was resolved on 2026-04-23 at 18:34 UTC. Cause: The failures were due to a temporary service outage on ServiceTitan’s end, disrupting sync processes. Resolution: The issue was resolved once ServiceTitan deployed a fix on their end

We have resolved this incident.

This incident has been resolved. We have observed that error rates have returned to normal levels, and the service is operating as expected. Incident Summary Description: We identified an issue affecting multiple Twitter Ads Connections Sync Failures Due to Access Level/Permission Error (Code 453) Timeline: The issue began on 2026-04-21 20:43 UTC and was resolved on 2026-04-22 02:23 UTC. Cause: X API started to fail report GZIP file authenticated download requests. Resolution: Authentication has been removed from GZIP file download requests.

We have resolved this incident. Incident Summary Description: We identified an issue affecting Apple App Store connections, which were failing with the error: “Unknown failure. This request can not be processed right now Reporter is currently unavailable.” Timeline: The issue began on April 20th at 22:45 UTC and was resolved on April 21st at 01:20 UTC. Cause: The sync failure was caused by an intermittent issue on the third-party side. Resolution: The issue was automatically resolved from the third-party side without any intervention from the Fivetran end.

This incident has been resolved. Incident Summary Description: We identified an issue where connectors using the Hybrid Deployment agent were unable to schedule new sync jobs. Timeline: This issue began on 2026-04-15 18:30 UTC and was resolved at 2026-04-16 23:20 UTC. Cause: The issue was caused by a recent code change. Resolution: The changes have been reverted, and the HD Agents affected will require a restart. During the restart, the latest update will be automatically applied, resolving the issue and preventing recurrence. Steps to restart your agent: For Docker and Podman deployments: ./hdagent.sh stop ./hdagent.sh start For Kubernetes deployments: 1. Get the HD Agent pod: kubectl get pod -n -l app.kubernetes.io/name=hd-agent 2. Kill the pod, which will force it to restart and pull the latest image kubectl delete pod

This incident has been resolved. We have observed that instance rates are returning to normal levels and affected connectors are syncing successfully. Incident Summary Description: We identified an issue for Twitter Ads which resulted in syncs failing with a "Null primary key found" error. Timeline: This issue began on 2026-04-14 at 16:30 UTC and was resolved at 23:30 UTC. Cause: Twitter API started returning null values in a primary key field (segment) for some reports, causing sync failures. Resolution: Twitter resolved the issue within the source: https://devcommunity.x.com/t/x-api-service-outage/262778 A hotfix was deployed by Fivetran to prevent further failures by skipping keys where the primary key was not present.

This incident has been resolved. Incident Summary Description: Twitter API started returning a Bad Authentication data error (code 215) while fetching successful async job results. Timeline: Impact start time around April 11th, 9:30 PM UTC to April 11th 01:57 UTC Cause: 3rd party error on Twitter side Resolution: We have implemented a workaround that skips the problematic endpoints for problematic accounts when this error occurs, and connectors are now back to normal.

Incident Summary. Description: Twitter API started returning Internal Server Error with code 500 when we try to fetch reports. Timeline: Impact start time around April 10th 21:54:44 UTC to April 11th 12:47:37 UTC Cause: 3rd party error on twitter side Resolution: We have implemented a workaround for this by skipping reports for problematic accounts and connectors are back to normal.

This incident has been resolved. Incident Summary Description: We identified an issue where dbt core and Quickstart transformation jobs are failing with "UnsatisfiedDependencyException" error. Timeline: The issue began on April 10, 2026 at 00:40 UTC and was resolved on April 10, 2026 at 01:30 UTC. Cause: The issue was caused by a recent code change made to the transformation. Resolution: Engineering identified the root cause and reverted the recent changes. After deploying the revert, the transformation resumed normal operation.

This incident has been resolved. We have observed that instance rates have returned to normal levels, and affected connectors are syncing successfully. Incident Summary Description: We identified an issue with Twitter Ads and Twitter Organic connections, which resulted in sync failures with an "HTTP 503 (Service Unavailable)" error. Timeline: This issue began on 2026-04-07 at 15:00 UTC and was resolved at 23:00 UTC. Cause: The Twitter Ads API '/stats/jobs/accounts/' endpoint started returning HTTP 503 (Service Unavailable) errors. Resolution: A hotfix was deployed to handle 503 errors, as well as to retry failed endpoints.

This incident has been resolved. We have observed that instance rates are returning to normal levels and affected connectors are syncing successfully. Incident Summary Description: We identified an issue for Amazon Selling partner connections which resulted in syncs failing with a "Cannot invoke "com.fivetran.platform.interfaces.connector.SharedContext.safeFileBufferFactory()" because "this.sharedContext" is null" error. Timeline: This issue began on April 8, 2026 at 11:00 AM UTC and was resolved on April 8, 2026 at 2:30 PM UTC. Cause: A bug has been introduced on our end due to which we started hitting context being null error . Resolution: A fix has been implemented on our end to mitigate the issue.