Serverless pools

Serverless pools was renamed - Databricks now calls it Standard and Dedicated access modes, as of July 2018.

Serverless pools → High Concurrency clusters → Shared / Single user access modes → Standard and Dedicated access modes

The original Cluster Type option for auto-managed compute shared by several concurrently active users, swapped out of the cluster dropdown in July 2018 for the High Concurrency cluster mode.

• It was the cluster type you could not tune - its replacement was announced as providing all the same benefits 'while also allowing flexibility in Spark and resource configuration', which is a polite way of saying serverless pools had none. 🔗
• The release note that retired it also renamed the dropdown it lived in, turning Cluster Type into Cluster Mode in the same breath. 🔗
• Databricks used the word serverless for a cluster type in 2018, years before serverless compute came back as a platform-wide idea - and one of the last things this one learned to do was run R. 🔗

Category

Compute / BI

In use from

June 2017🔗

Renamed

July 2018🔗

Announced at

Spark Summit 2017 (today's Data + AI Summit), as the first phase of Databricks Serverless🔗

Also known as

Serverless Pool, Serverless pool cluster type, Databricks Serverless

Verified

2026-09-27