4

I'm developing an ETL, whose extract step must read data from a table.

There are times that my user's access to that table is revoked. So, to avoid the ETL failing, before reading the table I must verify if I have permission to do it. If not, I'll just jump the extract step and run the rest of the ETL.

What's the proper way of verifying if my user has read permission over a table?

8

To check for a specific effective permission on a securable, use HAS_PERMS_BY_NAME*.

For example, to show if the current security context has SELECT permission on the Production.Product table in the AdventureWorks sample database:

SELECT 
    HAS_PERMS_BY_NAME
    (
        N'Production.Product', 
        N'OBJECT', 
        N'SELECT'
    );

Related: To list effective permissions on a securable, use sys.fn_my_permissions:

SELECT
    P.[entity_name],
    P.subentity_name,
    P.[permission_name]
FROM sys.fn_my_permissions
    (
        N'Production.Product', 
        N'OBJECT'
    ) AS P
ORDER BY
    P.subentity_name,
    P.[permission_name];

* The built-in function takes everything into account. SQL Server has to know how to correctly determine effective permissions. This function exposes that ability.

|improve this answer|||||

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.