Since the announcement of 12cR2 on-premises availability, the Oracle community become energetic and busy tweeting/blogging the new features, demonstrating installation & upgrades. Hence, I have decided to pick my favorite list of 12cR2 new features for Developers and & DBAs. Here is the high-level summary, until I write a detailed post for each feature ( excerpt from Oracle 12cR2 new features document).
: Pre 12cR2, this could be achieved through a workaround, now, the history command would do the magic for you.
: Materialized views can be used for query rewrite even if they are not fully synchronized with the base tables and are considered stale. Using materialized view logs for delta computation together with the stale materialized view, the database can compute the query and return correct results in real time.
For materialized views that can be used for query rewrite all of the time, with the accurate result being computed in real time, the result is optimized and fast query processing for best performance. This alleviates the stringent requirement of always having to have fresh materialized views for the best performance.
: In addition to ON COMMIT and ON DEMAND refresh, the materialized join views can be refreshed when a DML operation takes place, without the need to commit such a transaction. This is predominantly relevant for star schema deployments.
The new ON STATEMENT refresh capability provides more flexibility to the application developers to take advantage of the materialized view rewrite, especially for complex transactions involving multiple DML statements. It offers built-in refresh capabilities that can replace customer-written trigger-based solutions, simplifying an application while offering higher performance.
This new tool compares data blocks stored in an Oracle Data Guard primary database and its physical standby databases. Use this tool to find disk errors (such as lost write) that cannot be detected by other tools like the DBVERIFY utility.
: A subset standby enables users of Oracle Multitenant to designate a subset of the pluggable databases (PDBs) in a multitenant container database (CDB) for replication to a standby database.
This feature automatically synchronizes password files across Oracle Data Guard configurations. When the passwords of SYS, SYSDG, and so on, are changed, the password file at the primary database is updated and then the changes are propagated to all standby databases in the configuration.
: Currently, when a role change occurs and an Active Data Guard standby becomes the primary, all read-only user connections are disconnected and must reconnect, losing their state information. This feature enables a role change to occur without disconnecting the read-only user connections. Instead, the read-only user connections experience a pause while the state of the standby database is changed to primary. Read-only user connections that use a service designed to run in both the primary and physical standby roles are maintained. Users connected through a physical standby only role continue to be disconnected.
: The use of NOLOGGING for direct loads on a primary database has always been difficult to correct on an associated standby database. On a physical standby database the data blocks were marked unrecoverable and any SQL operation that tried to read them would return an error. Or, for a logical standby database, SQL apply would stop upon encountering the invalidation redo.
There is a new ROLLBACK parameter for the FINISH_REDEF_TABLE procedure that tracks DML on a newly redefined table so that changes can be easily synchronized with the original table using the SYNC_INTERIM_TABLE procedure.
The new V$ONLINE_REDEF view displays runtime information related to the current redefinition procedure being executed based on a redefinition session identifier.
: Nonpartitioned tables can be converted to partitioned tables online. Indexes are maintained as part of this operation and can be partitioned as well. The conversion has no impact on the ongoing DML operations.
: The partition maintenance operations SPLIT PARTITION and SPLIT SUBPARTITION can now be executed as online operations for heap organized tables, allowing the concurrent DML operations with the ongoing partition maintenance operation.
: Nonpartitioned tables can be moved as an online operation without blocking any concurrent DML operations. A table move operation now also supports automatic index maintenance as part of the move.
: Sharding with Oracle Database 12c Release 2 (12.2) is an architecture for suitable online transaction processing (OLTP) applications where data is horizontally partitioned across multiple discrete Oracle databases, called shards, which share no hardware or software. The collection of shards is presented to an application as a single logical Oracle database.
Stay tuned for Part 2..