ASM disk group

A disk group is a logical container for one or more ASM disks and is the highest level of data structure in ASM. When a database is configured to employ the disk group, the disk group then becomes the default location for its datafiles. The disk group
can be used to place various database file types, such as datafiles, online redo, archivelogs, RMAN backupsets, OCR and Voting disks (in 11g R2), and more. ASM also provides the flexibility of utilizing a single disk group by multiple ASM instances
and databases across a cluster.

After a disk group is successfully created and mounted for the first time in the ASM instance, the name of the disk group is automatically affiliated with the ASM_DISKGROUPS initialization parameter to be able to mount the disk group at ASM
instance restarts.

In general, when a datafile is created in a disk group, the datafile extents are striped/distributed evenly across the available disks of the disk group. Optionally, you can also set the following specified mirroring level at the disk group to protect the data
integrity by storing redundant copies of data (extents) in a separate failure group to cope with the disk outage symptom:

• External redundancy: Relies on the STORAGE (RAID)-level mirroring redundancy option to protect the data
• Normal redundancy: Provides a default two-way mirroring option
• High redundancy: Provides a three-way mirroring redundancy option of ASM files

As of 11g R2, the following limits have been imposed on the ASM instance:
• A maximum of 63 disk groups in a storage system
• 1 million files per disk group

The following diagram illustrates the structure of a disk group with three disks assigned to it: