Submit feedback on
Inefficient Use of Photon Engine in Databricks Compute
We've received your feedback.
Thanks for reaching out!
Oops! Something went wrong while submitting the form.
Close
Inefficient Use of Photon Engine in Databricks Compute
Josh Collier
Service Category
Compute
Cloud Provider
Databricks
Service Name
Databricks Clusters
Inefficiency Type
Inefficient Configuration
Explanation

Photon is enabled by default on many Databricks compute configurations. While it can accelerate certain SQL and DataFrame operations, its performance benefits are workload-specific and may not justify the increased DBU cost. Many pipelines, particularly ETL jobs or simpler Spark workloads, do not benefit materially from Photon but still incur the higher DBU multiplier. Disabling Photon by default and allowing it only where proven beneficial can reduce cost without degrading performance.

Relevant Billing Model

Databricks charges based on Databricks Units (DBUs), which are consumed per node per hour. Enabling Photon increases the DBU rate by up to 2.9x compared to standard compute. While Photon can significantly reduce runtime for certain workloads, in many environments the performance benefit is marginal, resulting in higher cost without proportional gain.

Detection
  • Identify compute clusters or jobs with Photon enabled
  • Review job configurations for SKUs or node types with Photon in the name
  • Check cluster configuration settings for the Photon toggle
  • Evaluate job-level performance data to determine if Photon meaningfully reduces runtime for the associated workload
Remediation
  • Update default compute configurations to disable Photon for general-purpose or low-complexity workloads
  • Restrict users from enabling Photon unless justified by benchmarked performance gains
  • Establish cluster policies or templates that exclude Photon by default and allow opt-in only under specific conditions
Relevant Documentation
Submit Feedback