What is new in 12c for Backup and Recovery?

OOW has announced the new Oracle Database 12c, and this new version of the database product as expected has introduced many new features and enhancements for Backup and Recovery, on this post I will introduce you to some of them.

Of course, I cannot start talking about 12c without talk first about a revolutionary whole new concept that was introduced with this new version, it is called, “Pluggable Database”.

Pluggable Database

We are now able to have multiple databases sharing a single instance and Oracle binaries, and each of the databases will have their own configuration and parameters and what is better, each database will be completely isolated of each other without either know that each other exists.

A Container Database (CDB) is a single physical database that contains a root container with the main Oracle data dictionary, and at least one Pluggable Database (PDB) with specific application data, a PDB is a portable container with its own data dictionary including metadata and internal links to the system supplied objects in the root container, and this PDB will appear to an Oracle Net client as a traditional Oracle database. The CDB also contain a PDB called SEED, that is used as a template when a empty PDB needs to be created.

When creating a database on 12c, you can now create a CDB with one or more PDBs, and what is even better; you can easily clone a PDB, or unplug it and plug it on a different server with a preinstalled CBD if your target server is running out of resources such as CPU or memory. Many years ago, the introduction of external storages gave us the possibility to store data on external devices and the flexibility to plug and unplug them to any system independent of their OS (Operating System). As per example, you can connect an external device to a system using Windows XP and read your data without any problems, later you can unplug it and connect it to a laptop running Windows 7 and you will still be able to read your data. Now with the introduction of Oracle Pluggable Databases, we will be able to something similar with Oracle when upgrading a PDB , making this process simple and easy. All you will need to do to upgrade a PDB as per example (see figure bellow) is to unplug your PDB (step 1) that is using a CDB running 12.1.0.1, copy the PDB to the destination location with a CDB that is using a later version like 12.2.0.1 (step 2), and plug the PDB to the CDB (step 3), and your PDB is now upgraded to 12.2.0.1.

This new concept is a great solution for database consolidation and is very useful for multi-tenant SaaS (Software as a Service) providers, improving resource utilization, manageability, integration and service management, and You can have many PDBs you want inside a single container (A CDB can contain a maximum of 250 PDBs), also a PDB is fully backward compatible to pre-12.1 database releases.

RMAN New Features and Enhancements

Now we can continue and take a fast and closer look on some of the new features and enhancements introduced in this database version for RMAN.

Container and Pluggable Database Backup & Restore

As we saw before, the introduction of 12c and the new Pluggable Database concept made possible to easily centralize multiple databases maintaining the individuality of each one when using a single instance. The introduction of this new concept also forced Oracle to introduce some new enhancements to the already existent BACKUP, RESTORE and RECOVERY commands, to enable us to be able to make an efficient backup or restore of the complete CDB including all PDBs, or just one of more PDBs or if you want to be more specific, you can also just backup or restore one or more tablespace from a PDB.

Enterprise Manager Database Express

The Oracle Enterprise Manager Database Console or Database Control that many of us used to manage an entire database its now deprecated and replaced by the new Oracle Enterprise Manager Database Express. This new tool uses flash technology and allows the DBA to easily manage configurations, storage, security and performance of a database. Note that RMAN, Data Pump and the Oracle Enterprise Manager Cloud Control are now the only tools able to perform backup and recovery operations in a Pluggable Database environment. In other words, you cannot manage Oracle Backups on Enterprise Manager Database Express.

Support for Third-party Snapshot

In the past when using a third-party snapshot technology to make a backup or clone of a database you was forced to change the database to backup mode (BEGIN BACKUP) before execute the storage snapshot. This requirement is now not necessary if the following conditions are met:

• The database crash is consistent at the point of the snapshot;
• Write ordering is preserved for each file within the snapshot;
• The snapshot stores the time at which the snapshot is completed.

Table Recovery

On previous versions of Oracle database, the process to recover a table to an specific point-in-time was never easy, but now Oracle solved this major issue introducing the possibility to do a point-in-time recovery of a table, group of tables or even table partitions without affect the remaining database objects using RMAN, making this process easy and faster than ever before. Remember that previously Oracle have introduced features like database point-in-time recovery (DBPITR), tablespace point-in-time recovery (TSPITR) and Flashback database, this is an evolution of the same technology and principles

The recovery of tables and table partitions is useful when in the following situations:

• To recover a very small set of tables to a particular point-in-time;
• To recover a tablespace that is not self-contained to a particular point-in-time, remember that TSPITR can only be used if the tablespace is self-contained;
• To recover tables that are corrupted or deleted with the PURGE option, so the FLASHBACK DROP functionality is not possible to be used;
• When logging for a Flashback table is enabled, but the flashback target time or SCN is beyond the available UNDO;
• To recover data that was lost after a data definition language (DDL) operation that changed the structure of a table.

Data Pump New Features and Enhancements

Now is the time to take a closer look on some of the new features and enhancements introduced in this database version on Data Pump.

Disabling Logging on Data Pump Import

A new feature on Data Pump introduced with 12c is the possibility to disable logging generation during an import operation, allowing us to have faster imports due that the redo log information is not written to disk or even archived. This is particularly useful for large data loads like database migrations.

Exporting Views as Tables

Another new feature introduced to Data Pump, is the option to export a view as a table. In this case, Data Pump will include in the dump file the corresponding table definition and all data that was visible in the view, instead to only write the view definition. This allows the Data Pump import to create it as a table with the same columns and data as the original view during the import process. All objects depending on the view will be also exported as if they were defined on the table, grants and constraints that are associated with the view will be now recorded as grants and constraints on the corresponding table in the dump file.

Extended Character Data Types

This new version of the Oracle database extends the maximum size of the VARCHAR2, NVARCHAR2 from 4000 bytes to 32767 bytes and the RAW data type from 2000 bytes to 32767 bytes. Columns with a declarated length of 4000 bytes or less will be stored inline and for columns with a length greater than 4000 bytes are called extended character data type columns and will be stored out-of-line, leveraging the Oracle Large Object (LOB) technology. Data Pump utilities and the related packages DBMS_DATAPUMP and DBMS_METADATA (PL/SQL) are now modified to support the extended data types.

I know that many more new features will come with Oracle 12c, but we will need to wait until the product is released to see what else was included!

REFERENCES

• Larry Ellison, CEO, Oracle: Oracle Open World Welcome Keynote: Oracle and Fujitsu
• Andrew Mendelsohn, Senior VP, Database Server Technologies, Oracle: General Session: What’s next for Oracle Databases?
• Bryn Llewellyn, Distinguished Product Manager, Oracle & Kumar Rajamani, Architect, Oracle: Consolidating Databases with Latest Generation of Database Technology
• Twitter and comments on OTN Forums.
• Computer World UK

DISCLAIMER

The above document is intended to outline Oracle’s general product direction based on the information compiled from the references listed. It is intended for information purposes only for learning. The development and release of these functions including the release dates remain at the sole discretion of Oracle and no documentation is available at the time of this writing. The command shown may or may not be accurate when the final database release goes GA. Please refer Oracle documentation when it becomes available.