admin管理员组文章数量:1389798
When I do some DML statements using a view "the_source_view" direcly in SQL as a basis, all works fine. I can do inserts in another Delta table based on the data coming out of this view. It works too from a Job: WorkFlow -> Task -> Notebook containing SQL statements.
Note: this view "the_source_view" is based on another MV created by a colleague of another team, on which we have at least the grant select privilege. But I'm not owner of this.
But the object is in the same Catalog and same schema.
Now if I try to execute the same statement via PySpark, like I would do as a dynamic SQL in Oracle, I get the following error:
[INSUFFICIENT_PERMISSIONS] Insufficient privileges: User does not have SELECT on Table 'mycatalog.mychema.__materialization_mat_xxyyzzsome technical_id_the_source_view_1'. SQLSTATE: 42501
I checked in the catalog, and this object mentioned "materialization_mat_xxyyzzsome technical_id_the_source_view_1" is appearing as a dependency of "the_source_view_1", as an Upstream Lineage.
What is missing there in terms of privileges? When I look at the permissions, I don't see any difference with other tables or views.
The PySpark execution mean that I'm using a All Purpose compute.
When I do some DML statements using a view "the_source_view" direcly in SQL as a basis, all works fine. I can do inserts in another Delta table based on the data coming out of this view. It works too from a Job: WorkFlow -> Task -> Notebook containing SQL statements.
Note: this view "the_source_view" is based on another MV created by a colleague of another team, on which we have at least the grant select privilege. But I'm not owner of this.
But the object is in the same Catalog and same schema.
Now if I try to execute the same statement via PySpark, like I would do as a dynamic SQL in Oracle, I get the following error:
[INSUFFICIENT_PERMISSIONS] Insufficient privileges: User does not have SELECT on Table 'mycatalog.mychema.__materialization_mat_xxyyzzsome technical_id_the_source_view_1'. SQLSTATE: 42501
I checked in the catalog, and this object mentioned "materialization_mat_xxyyzzsome technical_id_the_source_view_1" is appearing as a dependency of "the_source_view_1", as an Upstream Lineage.
What is missing there in terms of privileges? When I look at the permissions, I don't see any difference with other tables or views.
The PySpark execution mean that I'm using a All Purpose compute.
Share Improve this question edited Mar 13 at 17:58 isherwood 61.2k16 gold badges121 silver badges170 bronze badges asked Mar 13 at 15:00 SunchaserSunchaser 113 bronze badges 1- Please don't tag your titles. See How to Ask. – isherwood Commented Mar 13 at 17:58
1 Answer
Reset to default 0It seems that the solution is to use another setting on the "all-purpose" compute, Access Mode "Standard".
In this case, it worked.
本文标签：
版权声明：本文标题：databricks - PySpark throwing "Insufficient privileges" error while pure SQL code is working - Stack Overflow 内容由网友自发贡献，该文观点仅代表作者本人， 转载请联系作者并注明出处：http://www.betaflare.com/web/1744694600a2620201.html， 本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容，一经查实，本站将立刻删除。
发表评论