Hello -- I am building a Sentry Plugin for Presto. But was curious if one could simply leverage the HDFS-Sentry plugin and have that sync with the Hive Metastore ( https://www.cloudera.com/documentation/enterprise/latest/topics/sg_hdfs_sentry_sync.html) because Presto uses the objects defined in the Hive Metastore. This would seem like a decent work around until a Presto plugin is fully functioning.
It seems that Spark SQL works this way. Whereas Hive and Impala have their own plugins. So it seems I could simply do the same as Spark SQL for Presto. The disadvantages to using the HDFS-Sentry plugin would be that it likely does not support column access control nor could I view or manage the Sentry permissions from Presto. Which I'm assuming Spark SQL has the same limitations. Could someone confirm my understanding? Or correct me if I am wrong here? Thank you.
