Combines runnable code cells, in Python, SQL, Scala or R, with their results in one interactive Azure Databricks document; it runs on attached compute, and a job can run it as a task.
Also called Databricks notebook.
Read more: Microsoft Learn
In the Ultra Transcenders books
Each book explains Notebook in context, with comparison tables and the common traps.
Terms in this definition
- Serverless
Compute tier for single Azure SQL databases that scales automatically, pauses when idle and charges by the second. It is offered in General Purpose and Hyperscale, not Business Critical, and reserved capacity does not apply.
- Databricks
Analytics platform built on Apache Spark where data is processed in notebooks; Unity Catalog is the governance model it recommends.
- Agents (classic) API
First-generation Foundry Agent Service API, based on threads, messages and runs. It is deprecated, replaced by conversations and responses, and retires on 31 March 2027.
- Job
A sequence of steps executed together on one agent or runner, or on the server for agentless work. While running, each occupies one of your parallel jobs.
Related terms
- All-purpose compute
Interactive classic compute for notebook work, shareable between several people and restarted by hand. Its pricing differs from jobs compute, and production jobs are better run elsewhere.
- Best Practice Analyzer
A prebuilt notebook you can launch for any semantic model in the Power BI service. It scores your design on over 60 checks covering speed, DAX quality, avoiding errors, upkeep and formatting, and running it calls for Build rights plus a Fabric-backed workspace.
- Databricks Connect
Lets Spark code written in an IDE, a notebook server or an application run on remote Azure Databricks compute. Built on Spark Connect, the library can't be installed where PySpark is already installed locally.
- Default lakehouse
Pinned to a notebook, it determines where relative
Files/andTables/paths and Spark SQL table names without a qualifier point. If you switch to another lakehouse or rename it, the session must be restarted. - Driver node
In classic compute, the machine running the Spark driver: it keeps notebook state and collects results returned by
collect(), while workers host the executors. Unless you pick otherwise, its VM size matches the workers'. - Full mode
An environment library publishing option that works out dependencies up front, building a stable snapshot during a publish lasting around 3-6 minutes and deploying it as each session begins. That makes it the better fit for pipelines and scheduled jobs, compared with Quick mode, where packages get installed when the notebook session opens.
- High concurrency mode
A Fabric Spark mode where notebooks, or notebook activities in pipelines, run by one user with matching default lakehouse, Spark configuration and libraries reuse a single Spark session. They start more quickly and only the session that started it is charged.
- IPYNB
Jupyter's notebook format. Unlike source-format notebooks, notebooks in this format can be created and committed from Git folders with their cell outputs, if you want them.