Datasheet
Working with the ASM Instance
17
group attributes are set using the attribute clause of the create diskgroup and alter
diskgroup
commands. The following attributes can be set on a specific ASM disk group:
Au_size This is the disk group allocation unit (AU) size. The value defaults to 1MB and
can be set only when the disk group is created. You must modify the AU size of the disk
group if you want the disk group to be able to hold larger amounts of data. A disk group
with the default AU size will be able to grow to 35TB (normal redundancy). Increasing the
AU size will significantly increase the maximum size of the disk group. The maximum AU
size is 64MB.
Compatible.rdbms Indicates the database version that the disk group is compatible with
at a minimum (default is 10.1). This value should be equal to or greater than the compat-
ibility parameter of the database(s) accessing the ASM disk group. This value cannot be
rolled back once set.
Compatible.asm Indicates the ASM instance version that the disk group is compatible with at
a minimum (default is 10.1).
Compatible.asm must always be set to a value equal to or greater
than
compatible.rdbms. Once compatible.asm is set for a disk group, it can not be rolled
back to an earlier value.
Disk_repair_time Indicates the length of time that the disk resync process should maintain
change tracking before dropping an offline disk. The default for this parameter is 3.6 hours.
Disk group attributes can be viewed using the
V$ASM_ATTRIBUTE view. You can see some
examples of setting compatibility here:
Create diskgroup robert01 external redundancy
Disk ’/oracle/asm/ASM_DISKGROUP_robert01.asm’
Attribute ’ccompatible.asm’=’11.1.0’;
Alter diskgroup robert01 set attribute ’DISK_REPAIR_TIME’=’1200M’;
Alter diskgroup robert01 set attribute ’compatible.asm’=’11.1.0’;
ASM Fast Disk Resync
The redundancy features of ASM make it possible for an ASM disk group to survive the loss
of a disk associated with that disk group. Disk loss can result from a number of reasons, such
as loss of controller cards, cable failures, or power-supply errors. In many cases, the disk itself
is still intact. To allow for sufficient time to recover from disk failures that do not involve the
actual failure of a disk, ASM provides the ASM fast disk resync feature.
By default, when a disk in an ASM disk group fails (including any associated infrastructure
pieces), the disk will be taken offline automatically. The disk will be dropped some 3.6 hours
later. As a result, you have only 3.6 hours by default to respond to a disk outage. If you cor-
rect the problem and the physical disk media is not corrupted, then ASM fast disk resync will
quickly resynchronize the disk when it comes back online, correcting the problem very quickly.
This type of resynchronization is much faster than rebuilding a newly added disk should the
disk media be corrupted.
95134c01.indd 17 1/28/09 9:43:42 AM