You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In ranger-hive-plugin,view permissions and original table permissions are separated。
So you can create views on your table with specified conditions(specify partitions, rows, columns),then grant views privilege to some users but not the original table。
After actual verification and reading code,i find spark-authorizer do not support the feature。
Do you think this feature is necessary? In fact, I have already supported it by modifying
the code. If you think it is necessary, I am happy to share the code and ask for your guidance.
The text was updated successfully, but these errors were encountered:
Both spark and spark-authorizer need to be modified.
directions:
spark LogicalPlan will apply rule extensions “org.apache.spark.sql.catalyst.analysis.EliminateView” before apply rule “org.apache.spark.sql.catalyst.optimizer.RangerSparkAuthorizerExtension”。
“EliminateView” will remove the view information,so when “RangerSparkAuthorizerExtension” builder privileges,it can not get the view Information。In fact, "RangerSparkAuthorizerExtension" does not to get the view informations at all.
modified:
1, in spark,before apply rule “EliminateView”,we can store the view information
2, in spark-authorizer, we can get the view information when build privileges
In ranger-hive-plugin,view permissions and original table permissions are separated。
So you can create views on your table with specified conditions(specify partitions, rows, columns),then grant views privilege to some users but not the original table。
After actual verification and reading code,i find spark-authorizer do not support the feature。
Do you think this feature is necessary? In fact, I have already supported it by modifying
the code. If you think it is necessary, I am happy to share the code and ask for your guidance.
The text was updated successfully, but these errors were encountered: