Hi, We are using ranger to authorize HDFS requests. And we find an issue that if there is an HDFS path authorized to too many users (about hundreds of users), it will take a long time to update the authorization rule (up to 10 seconds more). After reading the source code, it seems that ranger admin will delete all the old rules, then insert newly configured rules to implement the update operation.
I'd like to ask if there is system design consideration here to use delete-and-insert pattern to implement update operation, and if there are some optimizations to help us accelerate the operation ? Thanks
