||Oracle Tips by Burleson
10g Grid Computing
Real Application Cluster Architecture
RAC Database Nodes
The next main component is the host or node
where the database instance resides. This is the place where
data-actual database processing takes place. RAC Database system
provides scalability and high availability. In order to provide
large database computing power and maintain large workload, you need
to extend the number of nodes. Most cluster frameworks today differ
in terms of the maximum number of nodes they can handle. Most can
support a minimum of 4 nodes today, and some can support hundreds.
This will also be driven by your scalability objectives and capacity
The nodes themselves may or may not need to be
scalable to provide additional capacity in terms of CPU or memory.
Unless one uses an expensive SMP server, scalability will be an
issue. The ability to scale both within a machine as well as across
machines is often desirable.
Oracle RAC can be implemented on a wide range
of servers from a clustered group of single CPU Windows boxes to a
cluster of 32-CPU SUN E10000 boxes. One of the more promising
architectures for RAC is blade architecture.
In blade architecture, the servers are inserted
into a pre-configured RAC similar to NIM (Nuclear Instrumentation
Modules) in a NIM bin. Rather than a horizontal orientation, they
have a vertical one. Blade servers are essentially self-contained
servers that rest on a single backplane. The backplane in a blade
array provides power, network, and management connections, reducing
cabling and overall component expense. Many blade architectures are
hot-pluggable, allowing online addition and removal of servers from
Linux machines are now able to scale up to 8
CPUs, but the majority of systems are 2 or 4 CPU nodes. SMP
scalability in Linux beyond 4 CPU’s is not well proven, so the
current Oracle recommendation is to stick with 4-CPU machines. The
Blade Servers provide low cost 4-way or 2-way servers for building
the large cluster environment.
At the same time, Oracle RAC can also run on
platforms that allow sub-setting of CPUs, such as the SUN E10000,
E15000, and the HP Superdome. In the case of CPU sub-setting, the
single server is divided into multiple nodes, each running an
instance of Oracle9i RAC.
The database resides within a server. The
server or host is an important component in the provision of the
data service. Any failure in the host system causes the database to
Clustered servers utilize two or more nodes,
essentially keeping the extra nodes as standby or sometimes as extra
computing power, as in the case of the RAC system. With the help of
the additional nodes, we ensure that the standby node can provide
the same database service to the user community. However, when in
standby, it loses the performance and scalability level for which it
Clustering servers assures the administrators
and the application users that at least one node is alive. A
cluster, in its most general form, comprises two or more
interconnected computers that are viewed and used as a single,
unified computing resource. By using multiple systems, the impact of
the failure of any individual system is kept low by passing the
failed system’s workload to the remaining members of the cluster.
The standby node becomes functional, or becomes
the primary host, when the failed host is unable to provide any host
services. When some of the internal components fail and the failure
is non-recoverable without intervention, the server is declared not
available or simply ‘failed’. This indicates that there is a lot of
scope for keeping the internal components safe or redundant.
Before losing the server and resorting to the
use of the clustered backup node, there are many things we can do to
keep the components from failing. Let us examine these methods,
which act as the first level of redundancy. Some people call it
‘high availability without clustering’. In contrast to clustering,
system availability can be improved without adding additional
There are many features or options that add
value to the redundancy at the server level. Taking advantage of
such features helps avoid failures, and avoids degraded cluster
performance in systems like the RAC system. These features address
different subsystems of the server, such as the memory and
processors. Redundant components such as fans, power supplies, and
adapters can also provide higher availability, particularly when
used with software that provides monitoring and alerting capability
to the system administrators.
To make the servers more reliable, we should
use high-reliability components and best-system practices.
The above text is
an excerpt from:
Oracle 10g Grid & Real Application
Computing with RAC
by Mike Ault, Madhu Tumma
can offer world-class remote Oracle database support at super-affordable prices.
Oracle DBA service provides 100% of the remote Oracle database administration
needs for your company, and includes 24x7 access to our staff of 100% OCP
Certified Oracle DBAs.
We require a 12 month service commitment and include the following services:
- Initial configuration review and problem identification
- Installation of Oracle statistics collection mechanisms and quarterly
database growth summaries
- Hourly monitoring of your Oracle database for pending problems
- Periodic performance analysis & identification of tuning activities
- Twenty Four hour Oracle emergency support
- Reporting and resolving all serious Oracle alert log messages
- Free use of the BC
- Quick response emergency support for production database outages
- Four hours of free remote DBA support every month. You can use these free
hours for any DBA activity, including database analysis, system design,
production migrations or personal mentoring.
information, please visit www.dba-oracle.com or email