What causes Warehouse Scaling / Multi-Cluster Timeout?

The multi-cluster warehouse reached its maximum cluster count and new queries could not be dispatched. Scaling mode is set to Economy (scale out less aggressively than Maximized), causing a delay before new clusters start. All clusters are processing long-running queries, so the queue is not clearing even as clusters are running. The STATEMENT_TIMEOUT_IN_SECONDS is too short for the expected queue wait plus execution time

How do I fix Warehouse Scaling / Multi-Cluster Timeout?

Increase the maximum cluster count for the warehouse: `ALTER WAREHOUSE SET MAX_CLUSTER_COUNT = `.. Switch scaling policy from Economy to Standard to scale out more aggressively: `ALTER WAREHOUSE SET SCALING_POLICY = STANDARD`.. Review query queue depth in Snowsight Query History — if the queue depth is consistently high, the warehouse is undersized for the workload.. Consider splitting workloads onto dedicated warehouses by user group or use case to avoid contention.. Increase `STATEMENT_TIMEOUT_IN_SECONDS` on the warehouse to give queries more time to wait for a cluster slot.

Medium severitycapacity

Power BI Refresh Error:
Warehouse Scaling / Multi-Cluster Timeout

What does this error mean?

A Snowflake multi-cluster warehouse could not scale out in time to handle a query or workload, causing the query to be queued beyond the query timeout threshold.

Common causes

1The multi-cluster warehouse reached its maximum cluster count and new queries could not be dispatched
2Scaling mode is set to Economy (scale out less aggressively than Maximized), causing a delay before new clusters start
3All clusters are processing long-running queries, so the queue is not clearing even as clusters are running
4The STATEMENT_TIMEOUT_IN_SECONDS is too short for the expected queue wait plus execution time

How to fix it

1Increase the maximum cluster count for the warehouse: `ALTER WAREHOUSE <wh> SET MAX_CLUSTER_COUNT = <n>`.
2Switch scaling policy from Economy to Standard to scale out more aggressively: `ALTER WAREHOUSE <wh> SET SCALING_POLICY = STANDARD`.
3Review query queue depth in Snowsight Query History — if the queue depth is consistently high, the warehouse is undersized for the workload.
4Consider splitting workloads onto dedicated warehouses by user group or use case to avoid contention.
5Increase `STATEMENT_TIMEOUT_IN_SECONDS` on the warehouse to give queries more time to wait for a cluster slot.

Frequently asked questions

Does adding more clusters always solve performance problems?

More clusters help with concurrency (many parallel queries), not with individual slow queries. If a single query is slow, increase the warehouse size (XS → S → M) rather than adding clusters.