Snowflake Outage History

Snowflake customers in the Azure West US 2 region experienced a major service disruption where they were unable to start, resume, or manage virtual warehouses, with warehouses appearing stuck in associated states. The incident was caused by a power outage at a third-party cloud platform's data center, which affected underlying infrastructure and caused increased latency and intermittent connectivity. The primary impact lasted approximately 3 hours (05:15 to 07:24 UTC on May 29, 2026), with full resolution confirmed after 15.8 hours once the cloud provider restored their infrastructure.

Snowflake experienced an issue with Cortex Code in Snowsight where responses would stop mid-execution after completing only one step, either halting without error messages or repeating without progress. The problem was caused by a recent Cortex Code update that introduced a code defect preventing requests from completing properly. Snowflake resolved the incident by rolling back the problematic update after 1.4 hours of service disruption.

Snowflake investigated a potential issue with their Data Cloud service that could have caused delays or failures when customers replicated data between accounts or regions. The incident lasted 1.3 hours before Snowflake determined the issue was limited to internal systems and did not actually affect customer replication operations. Replication services were confirmed to be functioning normally throughout the incident.

Snowflake experienced an 8.3-hour outage in the Azure West US 2 region where customers observed stale or delayed data in account usage views and organization usage views, with usage reporting not reflecting activity after 16:00 UTC on May 4, 2026. The issue was caused by corrupted internal metering files that failed to read properly, blocking downstream processing for account usage updates. Snowflake resolved the incident by excluding the affected files and allowing processing to resume with backfill to catch up on delayed data.

Snowflake experienced a major service incident in the Azure UK South (London) region lasting 1.4 hours, caused by an unhealthy DNS processing component. Customers were unable to access Snowflake Openflow runtimes and experienced connectivity issues with Snowpark Container Services, including failed data pulls and pushes through Openflow Connectors. The issue was resolved by restarting the affected infrastructure and monitoring the environment to confirm service restoration.

Snowflake experienced a major outage in the GCP US Central 1 (Iowa) region from 12:25 to 13:45 UTC on April 16, 2026, lasting approximately 1 hour. Some application instances entered a degraded state and couldn't handle downstream connections, preventing customers in the affected region from accessing Snowflake services and features through Snowsight. The issue was resolved by restarting the affected instances, and service was fully restored after monitoring confirmed proper functionality.

Snowflake experienced a major service disruption in the Azure East US 2 region lasting 2.6 hours, where customers were unable to sign in, execute queries, manage data, or access Snowsight due to degraded metadata database infrastructure and Snowsight infrastructure issues. The incident was resolved by rerouting traffic away from the affected infrastructure to healthy systems, with some users experiencing temporary delays in query history during recovery.

A recent Snowsight UI update introduced a code issue that prevented customers from seeing or managing their Native Apps within the "Apps" page of the Snowsight interface. App execution and all other Snowflake features remained unaffected during the 75-hour incident. The issue was resolved by deploying a fix across all regions, with the final GCP - US East 4 region experiencing delays due to infrastructure issues that were eventually overcome.

Snowflake experienced a major outage in the AWS US West (Oregon) region lasting approximately 1.2 hours on March 21, 2026, caused by degraded infrastructure responsible for traffic routing and name resolution. Customers were unable to access core Snowflake services including sign-in, query execution, and data management, encountering HTTP 404 errors when attempting to use the Data Warehouse, Snowpipe, and Replication features. The issue was resolved after implementing a fix and monitoring confirmed full service restoration.

Snowflake experienced a 7.1-hour incident where customers in specified regions were unable to create or alter Snowflake alerts due to a code issue in a recent improvement to the alerts functionality. The issue was identified as stemming from faulty code in an alerts system update, and service was restored after implementing and rolling out a fix across the affected regions.