Azure / Azure/azure-data-lake-store-python

What's the permission needed for remove_acl_entries()? The action fails with a service principal, even it already has RWX permission, also the storage owner

Open
#331 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
76
Forks
79
PR merge metrics
No merged PRs in 30d

Description

### Description
Try to authenticate with a service principal in the codes. The service principal has already been given the owner IAM role of the ADLS gen1 storage. Also, it has RWX permission on the folder/files.

But it still fails to remove the ACL entries with the remove_acl_entries() function.
Traceback (most recent call last):
File "C:\Users\xinrantao\PycharmProjects\pythonProject\acl_remove_test.py", line 42, in
main()
File "C:\Users\xinrantao\PycharmProjects\pythonProject\acl_remove_test.py", line 38, in main
adl.remove_acl_entries("/testfolder1", "user:518b91c4-a815-409d-8915-ad6249f5095d", recursive=False, number_of_sub_process=None)
File "C:\Users\xinrantao\AppData\Local\Programs\Python\Python39\lib\site-packages\azure\datalake\store\core.py", line 534, in remove_acl_entries
self._acl_call('REMOVEACLENTRIES', path, acl_spec, invalidate_cache=True)
File "C:\Users\xinrantao\AppData\Local\Programs\Python\Python39\lib\site-packages\azure\datalake\store\core.py", line 459, in _acl_call
to_return = self.azure.call(action, posix_path, **parms)
File "C:\Users\xinrantao\AppData\Local\Programs\Python\Python39\lib\site-packages\azure\datalake\store\lib.py", line 450, in call
self.log_response_and_raise(response, PermissionError(path), level=exception_log_level)
File "C:\Users\xinrantao\AppData\Local\Programs\Python\Python39\lib\site-packages\azure\datalake\store\lib.py", line 358, in log_response_and_raise
raise exception
PermissionError: testfolder1

### Reproduction Steps
1. create the service principal, give it the owner IAM role of the ADLS gen1 storage:
![image](https://github.com/Azure/azure-data-lake-store-python/assets/92084556/6bfa8b36-e7af-45c6-a737-2a947e1b760f)

2. It has the RWX permission to the folder, which needs to remove the ACL entries:
![image](https://github.com/Azure/azure-data-lake-store-python/assets/92084556/b1ac1f00-2943-46b8-a6e3-2c790bb13f01)

3. Use the service principal as the authentication, try to remove the ACL entries of this folder from one of the user:

```
import azure.datalake.store
from azure.datalake.store import lib, core

def authenticate():
RESOURCE = "https://datalake.azure.net/"
tenant = "72f988bf-86f1-41af-91ab-2d7cd011db47"
client_id = "a8e33ec1-afae-4bfa-8a00-1de7b1f74a5f"
client_secret = "xxxxxxxxxxxxxxxxx"

adlCreds = lib.auth(tenant_id=tenant,
client_secret=client_secret,
client_id=client_id,
resource=RESOURCE)
adl = core.AzureDLFileSystem(adlCreds, store_name="xinranadlgen1")
return adl

def remove(path,acl_spec):
adl.remove_acl_entries(path, acl_spec, recursive=True, number_of_sub_process=None)

def version_check():
datalake_store_version = azure.datalake.store.__version__
print("Azure Data Lake Store Version:", datalake_store_version)

def main():
version_check()

adl = authenticate()
print("check before the removal:\n ******************\n", adl.get_acl_status("/testfolder1"))

print("\n ******************\n try with recursive=False: ")
adl.remove_acl_entries("/testfolder1", "user:518b91c4-a815-409d-8915-ad6249f5095d", recursive=False, number_of_sub_process=None)
```

4. it's interesting to find that if giving the owner of this specific folder to this service principal, then it seems to work.

Could you help check what's the minimum permission needed for this action? Is the owner of the folder/file necessary?
If it's necessary, how to give the owner of all the folder/files in a quick way? Because currently it relies on the service principal as authentication to remove all the ACL entries automatically. Thanks!

### Environment summary

**SDK Version:** What version of the SDK are you using? (pip show azure-datalake-store)
Answer here: 0.0.53

**Python Version:** What Python version are you using? Is it 64-bit or 32-bit?
Answer here: Python 3.9, 64-bit

**OS Version:** What OS and version are you using?
Answer here: Windows 11

**Shell Type:** What shell are you using? (e.g. bash, cmd.exe, Bash on Windows)
Answer here: PyCharm

Contributor guide

Open the contributing guide

Research direction

Start at the remove_acl_entries() and _acl_call() entry points shown in the traceback, then compare the service-principal setup with get_acl_status() and the documented ADLS Gen1 permission behavior. Use the supplied reproduction with SDK 0.0.53 to determine the minimum permission required and whether folder or file ownership is necessary; done means the requirement and any supported bulk approach are documented.

Written by the indexing model from the issue text.

Assessment

Tech stack
azure, python
Domain
authorization, cloud
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
38/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.