What causes Query Queued (Concurrency Limit)?

Warehouse MAX_CONCURRENCY_LEVEL is set to a low value (default is 8 simultaneous queries per cluster). Many concurrent users or scheduled jobs submitting queries at the same time. Long-running queries holding slots while new queries pile up in the queue. A single-cluster warehouse under heavy mixed workload (BI + ETL + ad-hoc queries). Multi-cluster warehouse scaling is disabled or the max cluster count is set too low

How do I fix Query Queued (Concurrency Limit)?

Enable multi-cluster warehouse scaling: ALTER WAREHOUSE my_wh SET MIN_CLUSTER_COUNT = 1, MAX_CLUSTER_COUNT = 3. Increase MAX_CONCURRENCY_LEVEL: ALTER WAREHOUSE my_wh SET MAX_CONCURRENCY_LEVEL = 16. Separate BI and ETL workloads onto dedicated warehouses to prevent contention. Identify long-running queries holding slots: SELECT * FROM TABLE(INFORMATION_SCHEMA.QUERY_HISTORY()) WHERE EXECUTION_STATUS = 'RUNNING' ORDER BY TOTAL_ELAPSED_TIME DESC. Set query timeouts to prevent long queries from monopolizing slots indefinitely

Medium severityperformance

Power BI Refresh Error:
Query Queued (Concurrency Limit)

What does this error mean?

Queries are queuing rather than executing immediately because the virtual warehouse has reached its maximum concurrency limit. Queued queries wait for running queries to complete before slots become available.

Common causes

1Warehouse MAX_CONCURRENCY_LEVEL is set to a low value (default is 8 simultaneous queries per cluster)
2Many concurrent users or scheduled jobs submitting queries at the same time
3Long-running queries holding slots while new queries pile up in the queue
4A single-cluster warehouse under heavy mixed workload (BI + ETL + ad-hoc queries)
5Multi-cluster warehouse scaling is disabled or the max cluster count is set too low

How to fix it

1Enable multi-cluster warehouse scaling: ALTER WAREHOUSE my_wh SET MIN_CLUSTER_COUNT = 1, MAX_CLUSTER_COUNT = 3
2Increase MAX_CONCURRENCY_LEVEL: ALTER WAREHOUSE my_wh SET MAX_CONCURRENCY_LEVEL = 16
3Separate BI and ETL workloads onto dedicated warehouses to prevent contention
4Identify long-running queries holding slots: SELECT * FROM TABLE(INFORMATION_SCHEMA.QUERY_HISTORY()) WHERE EXECUTION_STATUS = 'RUNNING' ORDER BY TOTAL_ELAPSED_TIME DESC
5Set query timeouts to prevent long queries from monopolizing slots indefinitely

Frequently asked questions

Does multi-cluster warehouse scaling add cost?

Yes — each additional cluster runs at the full warehouse credit rate. Set a MAX_CLUSTER_COUNT that balances query performance against credit budget, and use economy scaling policy to minimize cost.

Can I see the current queue depth?

Yes: SELECT * FROM TABLE(INFORMATION_SCHEMA.QUERY_HISTORY()) WHERE QUERY_TYPE = 'SELECT' AND EXECUTION_STATUS = 'QUEUED'. The QUEUED_PROVISIONING_TIME column shows how long each query waited.