Enable secure data sharing across organizations using Delta Sharing, an open source protocol that requires no ETL or vendor lock-in. This tutorial covers the complete workflow: creating a share, adding recipients, granting permissions, consuming shared data with the Delta Sharing client, and revoking access.
Source: official Databricks documentation · Databricks announcement
Key takeaways
- Delta Sharing is an open source protocol enabling secure cross-organization data sharing without ETL or vendor lock-in.
- Credential files are downloaded once from activation links and used by consumers to access shared tables.
- Permissions are granted with SQL GRANT commands specifying exact table access and access can be revoked at any time.
- Consumers install Delta Sharing library and load shared data as pandas or PySpark DataFrames in their environments.
- Delta Sharing has connectors available for various tools and supports read-only access by default.
Chapters
- 0:00 — Secure data sharing with Delta Sharing
- 0:32 — Provider side: create a share and add tables
- 1:04 — Add recipients and download credentials
- 1:34 — Grant permissions to recipients
- 2:07 — Consumer side: install the Delta Sharing library
- 2:38 — Load and query shared data
- 3:08 — Revoking access and permissions management
- 3:39 — UI walkthrough and Delta Sharing connectors
Frequently asked questions
How do I create and share a table with Delta Sharing?
Create a share with CREATE SHARE, add Delta tables to it, create a recipient to get an activation link, then grant SELECT permissions with GRANT commands. The recipient downloads credentials once and can access the shared data.
Can I revoke access to shared data?
Yes, you can use REVOKE commands to remove permissions. Once revoked, consumers will get permission denied errors when trying to access the data.
What formats can I consume shared data in?
After installing the Delta Sharing library, you can load shared data as pandas DataFrames or PySpark DataFrames depending on your environment.
Is Delta Sharing open source?
Yes, Delta Sharing is an open source protocol, making it vendor-agnostic and enabling integration with various third-party tools.
Full transcript
Auto-generated from the video and lightly edited for readability. Timestamps link to the moment in the video.
Show full transcript (13 segments)
0:00hello my name is Yousef and today I will show you how to use Dela sharing between a databas environment and your own LTAP so first as you can see I of course everything I will show you right now can be done through uh the API but also through the UI and the first thing I will do I'll create my first share so
0:25create share if not exist I will call it data sharing and I did add the com that is my first share and now I can show you like that inside this share I have nothing which makes sense because I just created the share and now what I will do I will just add inside this share my first Delta table so that I already have and of
0:47course I need to have the permissions to be able to add this table to this specific share and now I need to create the recipient which means the the consumer who will use this specific uh table so I'll call it call this person consumers and now I'll have here what we call an activation link so I need to go
1:12I'll copy paste this link as you can see it will create I it will lead me to a web page where I will download the credential file and just keep in mind that this credential file can only be downloaded uh once this is very very important and now that I have downloaded this uh
1:34file I need to Grant the permission to uh this recipient to access my data sharing so I will show grants of course if I show the grants of course I have no grants there so I need to Grant the select and I will be read only so I will grant select on my share Delta sharing to the recipient
1:57consumer so now I need to Grant the select and now that I did Grant the uh the select I'll go to my uh to my specific environment so I will open the python and now what I will do I will just of course I need to install uh first of all uh the Dela sharing Library
2:19so need to do pipe of course install Delta uh Delta sharing and once uh this library is installed I can go and copy paste my code where I will import my data sharing I will add
2:41my uh my share my my the configuration file I did download then of course I will list all the tables I will add the basic information and then I will load this data as pandas but of course if I have p park installed in my LTAP I can also use uh uh of course uh spark so if I print this data now you
3:06see that I have access to this data so I can run basic Transformations and store this data in in my LTAP if I want of course that's only uh a small data set and something very very important you can also rever the access whenever you want like here I will just revoke the access so if now I go back and try to
3:28load this data I will have ER tell me permissions the permission is denied and of course everything I did here in the UI in the API can also do it through the UI just need to go to data sharing and here you can create of course what we call uh create share I will call it
3:51sharing sharing data of course inside the share then what I can do I can add tables like I can go to demo use like both and here for example I can bring I don't know uh Titanic V2 for example and I can save this save this share and can I can also go if I want and I add I can add a recipient I can
4:14create a new recipient I can share it with data environment as I can share it with another uh person of course we have several connectors with Delta shage which works with uh for example uh pi and other other that you can find in Delta sharing and of course keep keep in mind that Delta sharing is an open source uh an open
4:38source protocol thank you for attending this session
Official Databricks resources
- Delta Sharing documentation — docs
- Databricks Free Edition — product
- Introducing Delta Sharing: an open protocol for secure data sharing (Databricks blog) — blog
Related
- Delta Lake 4.0: Connect, Clustering, Variant Type, Type Widening
- Mastering Delta Lake: Liquid Clustering, Time Travel & Vacuuming
- Liquid Clustering & Delta Lake Optimization: Author's Best Practices
- Delta Sharing Behind Firewalls: Private Endpoints Proxy Serverless NCC
- All Databricks video guides
- All tutorials & videos
Databricks Answers
Looking for a quick, source-backed answer? Browse Databricks Answers on data sharing, Marketplace & Clean Rooms — direct, documentation-validated answers to common Databricks questions, curated by Youssef Mrini, Senior Developer Advocate at Databricks.