MetricSign
EN|NLRequest Access
Low severityjob

Power BI Refresh Error:
QUEUED

What does this error mean?

A Databricks job run has been accepted but is waiting to start. The run is queued because no cluster is available yet, the maximum concurrent runs limit has been reached, or the job is waiting for a cluster from an instance pool.

Common causes

  • 1The job's maximum concurrent runs limit is set to 1 and a previous run is still in progress
  • 2A new cluster is being provisioned and the job is waiting for it to start
  • 3An instance pool used by the cluster has no available instances and is waiting for cloud provider allocation
  • 4Multiple jobs triggered simultaneously are competing for a limited number of shared clusters
  • 5A Databricks serverless compute queue is temporarily backed up

How to fix it

  1. 1Check whether a previous run of the same job is still in progress — this is the most common cause
  2. 2Review the job concurrency settings and increase the maximum concurrent runs if parallel execution is safe
  3. 3Check the instance pool status if the job uses a pool — the pool may need more capacity
  4. 4Monitor queue time trends — persistent queuing indicates a need for more compute resources
  5. 5For time-sensitive jobs, use pre-warmed clusters or instance pools with idle instances to minimize queue time

Frequently asked questions

How long will a run stay in QUEUED state?

Indefinitely, until a cluster becomes available or the run is canceled. There is no automatic timeout for QUEUED state by default.

Can I see how long a job waited in queue?

Yes — the Databricks run detail shows the queued time, setup time, and execution time separately so you can measure the delay.

Other job errors