We have defined a view layer on top of all tables in our edw. For example, every tabledatabase.table_t has a corresponding view defined as viewdatabase.table. We have a semantic view layer where we have tried to virtualize some of our dimension and fact tables. When defining complex views in the semantic layer, with a lock row for access, is it best to reference the tables (tabledatabase.table_t) or the access views of those tables (viewdatabasename.table)? If we define a complex view with a lock row for access and then reference the tables instead of their access views in the body, will we encounter any potential blocking issues? I know it is best practice to (1) define a one to one access view for every table and (2) always access the tables through these views.... Is this true when defining a complex view as described above? Will the lock row for access defined at the complex view level, place a lock row for access on every table referenced in the complex view?