Databases Reference
In-Depth Information
Stored data is automatically rebalanced when disks are added or removed. When a drive
fails, remirroring to remaining drives is automatic. These features make ASM ideal for
managing a database storage grid and allow you to use cheaper disk systems while
obtaining higher levels of availability and performance. Fast mirror resynchronization,
supported since Oracle Database 11
g
, enables faster recovery from transient failures
since ASM will only resynchronize changed ASM disk extents for limited duration
failures.
Oracle Database 12
c
introduces
Flex ASM
enabling ASM servers to be run on a separate
physical server from the Oracle Database servers. Large clusters of ASM instances can
be configured to support large numbers of database clients while reducing the overall
ASM footprint that would be present if Flex ASM were not deployed.
ASM and RAID Levels
Given the popular adoption of ASM and SAME for managing Oracle Database storage
on Oracle's engineered systems such as the Oracle Exadata Database Machine and other
Oracle and non-Oracle server and storage platforms, RAID 0+1 has become the most
common choice for deployment. For example, when Oracle installs the Exadata Data‐
base Machine, it includes full mirroring and striping as part of the standard installation.
Advanced compression techniques such as Hybrid Columnar Compression more than
offset additional storage required to deploy in this configuration.
Site and Computer Server Failover
The Oracle Database will recover automatically from a system crash. Automatic recov‐
ery protects data integrity, critical in a relational database, but also causes some down‐
time as the Oracle Database recovers. When a hardware failure occurs, the ability to
quickly detect a system crash and initiate recovery is crucial to minimizing the associated
downtime.
When an individual server fails, the instance(s) running on that server fail as well.
Depending on the cause, the failed server may not return to service quickly or be noticed
immediately. Either way, companies that wish to protect their systems from the failure
of a server typically employ multiple servers to achieve
failover
. Although failover
doesn't directly address the issue of the reliability of the underlying hardware, automated
failover reduces downtime from hardware failure. The primary means to mitigate un‐
planned downtime caused by server failures is through the use of Data Guard or Real
Application Clusters (RAC). Data Guard is also useful for protecting against site failures
while RAC is noteworthy for also providing timely database instance failure recovery.
We'll cover each of these in subsequent sections.
