Oracle 12c New Features - Expert Consolidation in Oracle Database 12c - page 42

You can see from Figure 2-2 that the RAC database is concurrently mounted by all instances. It is supported to
use a third-party file system for this purpose. For quite some time now Oracle has tried to deliver all the required
technology to run a clustered system, and its chosen technology—ASM—has matured and is more than adequate for
the purpose.
Usually access to shared storage is provided by Fiber Channel Host Base Adapters; however, other possibilities
exist as well. The inter-node communication mentioned earlier in this chapter is made possible by a dedicated,
cluster-private network. Most sites employ either multiple aggregated Gbit Ethernet ports or a couple of 10 Gbit
Ethernet ports for the purpose. More sophisticated deployments such as Exadata make use of Infiniband for the
cluster interconnect.
Once the operating system is configured for RAC the database administrator begins the installation of Grid
Infrastructure which is a mandatory component when installing RAC.
The final step is to install the database binaries over the Grid Infrastructure layer. The installer will detect that you
are trying to deploy the database binaries on a cluster and offer you a list of nodes to deploy the binaries to.
Brief Introduction to Automatic Storage Management
Automatic Storage Management (ASM) is a feature that was introduced in Oracle 10g Release 1. Among the primary
design goals is the intention to simplify storage administration by automating disk and file management tasks. As
with many new features introduced with the Oracle database, another aim is to reduce management overhead and
deployment costs. ASM is a generic alternative to a clustered file system (CFS)/cluster logical volume manager
that works on all supported platforms. The implementation of ASM differs greatly from other file systems which
you normally install on the operating system in form of kernel loadable modules and a user-mode part for the
administration. ASM on the other hand is an additional lightweight Oracle instance on each server. ASM provides
similar functionality to a classic CFS but includes volume management capabilities, such as disk balancing and
redundancy.
With the introduction of Oracle 10.2, ASM has become a popular choice for storage management, especially for
RAC. As a sign of its maturity, documentation and functionality have increased with each release since its inception,
and while initially documentation consisted of a chapter in the Administrator's Guide in 10g, it has now got a manual
of its own.
ASM is built on Oracle Managed Files (OMF), a feature that was first introduced in Oracle 9i. OMF greatly
simplifies the management of database files. Oracle ASM is by no means restricted to RAC as its position in this
chapter suggests. ASM is a great choice if you are looking for standardization of the storage layer.
Conceptually, ASM resembles the classical logical volume manager (LVM) known from Linux and AIX, but it
can do a lot more and—most importantly—it is cluster aware. Individual physical volumes, called ASM disks, are
aggregated into volume groups. Instead of volume groups these entities are referred to as ASM disk groups. In a RAC
environment, the ASM disks have to be located on shared storage accessible from all nodes in the cluster. Just like all
other cluster resources, disk groups are managed internally via Clusterware. Unlike for example Linux LVM, there
are no logical volumes created on top of a volume group. Instead, the disk group as a whole can be used to create
and store database files. Following the logic of the Oracle Managed File, all you need to do is define the disk group on
which your database should be created.
One of the striking beauties of Oracle ASM is the online-almost-everything approach. Adding and removing ASM
disks from an ASM Disk Group is an online operation. After an addition or removal of a disk, a so-called rebalance
operation ensues. Following a “Stripe-And-Mirror-Everything” (SAME) principle Oracle tries very hard to evenly
distribute information on ASM disks. Figure 2-3 compares the classic LVM approach with ASM.