WorkingWithFiles.md

Working with files

There are a number of ways to work with file and file-like interfaces in Databricks®. This can be confusing but is well documented in the included API level documentation and by Databricks.

• Files
• Workspaces
• Unity Catalog
• DBFS
  • DBFS API
• Standard/Shared clusters & storage

Furthermore the Spark™ API accessible via Databricks connect also enable file access. MATLAB® has built in support for working directly with cloud based object storage i.e. in S3 or Azure® Blob see: https://docs.databricks.com/api/workspace/workspace

Storage related FAQ issues

The following entries appear in this package's FAQ:

• Saving a .mat file to /dbfs fails
• Spark Submit fails to find jar on dbfs
• Azure Credential Passthrough name resolution in azPTAuth
• No FileSystem for scheme: abfss

Databricks references

Overview (recommended)

• https://docs.databricks.com/en/files/index.html

APIs

• https://docs.databricks.com/api/workspace/workspace
• https://docs.databricks.com/api/workspace/files
• https://docs.databricks.com/api/workspace/dbfs