A recent software update caused a major disruption to Snowflake’s cloud data platform, taking services offline across 10 of its 23 global regions for roughly 13 hours on December 16. During the outage, customers were unable to run queries or ingest new data, effectively halting analytics and data-processing workflows for many organizations relying on the platform.
According to Snowflake’s incident report, users attempting to access their data warehouses encountered “SQL execution internal error” messages. Beyond query failures, critical ingestion services such as Snowpipe and Snowpipe Streaming were also impacted, while data clustering operations showed signs of instability. These issues compounded the disruption, making it difficult for customers to maintain normal data operations throughout the incident.
Snowflake said its investigation traced the root cause to a backward-incompatible database schema change introduced in the latest release. The update caused older release packages to reference newly added schema fields, resulting in version mismatches. These conflicts led to failed operations and, in some cases, severely degraded performance as systems struggled to complete tasks.
The outage affected customers across multiple cloud providers and geographies, including regions hosted on AWS, Microsoft Azure, and Google Cloud. Impacted locations included parts of the United States, Europe, and Asia-Pacific, highlighting the broad reach of the incident. Snowflake has since worked to restore services and address the compatibility issue, but the event underscores the risks associated with schema changes in large-scale, multi-cloud environments.