Libraries added to a compute resource so all its notebooks and jobs share them. Because jobs can drift out of step this way, %pip or an environment is the better choice.
Also called cluster libraries.
Read more: Databricks docs
In the Ultra Transcenders books
Each book explains Compute-scoped libraries in context, with comparison tables and the common traps.
Terms in this definition
- ALL
A DAX function that ignores any filters and gives back every row of a table or every value of the named columns. Used within CALCULATE, it works as a modifier that clears filters, although REMOVEFILTERS states that intent more clearly where it is available.
- Share
In OpenSharing, the container you fill with tables, views, volumes, models and notebooks for recipients. You need
CREATE SHAREon the metastore to make one; add a schema and all its assets, including later ones, go with history. - Azure Machine Learning environment
Versioned asset that pairs a Docker image with a pip or conda specification, so every job or deployment using it gets identical dependencies. You point to it by name plus a version number, or by name with @latest.
Related terms
- CRAN
The R package repository, one of the places compute-scoped libraries can come from. Packages from CRAN can't be used on compute in standard access mode.
- Environments dependency mode
In Beta, classic compute using standard access can take its dependencies from a base environment. Docker, init scripts, cluster libraries and Spark environment variables can't be used alongside it.