Databricks Repos API Limitations

[Ha2001]

Hi, I have started using databricks recently, and I'm not able find a right solution in the documentations. i have linked multiple repos in my databricks workspace in the repos folders, and I wanted to update the repos with remote AzureDevops repositories parallelly. Is there any limitations to the REPOs API such as the number of update requests it can handle in a workspace simultaneously?

[Kaniz_Fatma]

Hi @Ha2001, 

• In a Databricks repo, the total number of all files should not exceed 10,000.

1. • Similarly, the total number of notebooks should not exceed 5,000.
   • For any Git operation (such as cloning, committing, or pushing), memory usage is limited to 2 GB.
   • Disk writes during these operations are capped at 4 GB.
   • If you attempt to clone a Git repo that is 5 GB in current size, you might encounter a failure due to these limits.
   •  If you’re dealing with a large number of repos or need to perform parallel updates, consider distrib....

[Ayushi_Suthar]

Hi @Ha2001 , Good Day! 

Databricks API has a limit of 10 per second for the /repos/* combined requests in the workspace. You can check the below documentation for the API limit: 

https://docs.databricks.com/en/resources/limits.html#:~:text=Git%20folders%20API,Yes

For other limitations on the Repos, you can check the below document: 

https://docs.databricks.com/en/repos/limits.html

Please let me know if this helps and leave a like if this information is useful, followups are appreciated.
Kudos
Ayushi