SFRAC

Hello Auspexian,

VCS provides high availability of resources configured in cluster. e.g  shared diskgroups, mout points, IP address, application etc. So that, if one node detects any fault on any resource or system goes down, the applications can fail over to another node and provide maximum uptime.

Oracle RAC is for providing performance aid on oracle application with shared storage and multiple instances of oracle on multiple nodes with centralised management.

Yes,  Veritas volume manager provides redundancy with disk redundancy.

Please check :
http://sfdoccentral.symantec.com/sf/5.1/linux/pdf/sfcfsrac_notes.pdf
http://sfdoccentral.symantec.com/sf/5.1/linux/pdf/vxvm_admin.pdf

Regards,
~Anoop

Hi Auspexian....

Also, SFRAC is a resilient infrastructure to run Oracle RAC inside Veritas cluster. Some key points here would be:

a) RAC controlled with in VCS, so you would configure in such a way that Oracle CRS would be controlled by VCS.

CRS is basically a process & I believe the failure detection (split brain of cluster) is far enhanced in veritas cluster since it uses kernel modules & are realtime. LLT/GAB/IOFENCING

b) Oracle Disk Manager module is gives with SFRAC which helps you to attain same level of performance as you would get directly with RAC...

To my experience, SFRAC would be a resilient solution...

Gaurav

Thank you everyone.

Is it possible to use Oracle RAC with just Volume Manager and Veritas Clusterd Filesystem?
without VCS?

Our DBA team is pushing back on VCS as they are not familiar with it. They said RAC can do everything, why addon VCS :-)

Tks

J.

Hi Again,

Veritas clustered filesystem will again install veritas cluster ..... so no escape here....

However it is possible to use just veritas volume manager with RAC using CRS for cluster....



Gaurav

We are thinking Clustered Volume Manager+DMP would give us the SAN management and Oracle RAC on top. Does CVM use VCS?

Tks

J.

From SFRAC Install guide, some features why SFRAC could prove better over plain RAC

Veritas Storage Foundation for Oracle RAC (SF Oracle RAC) from Symantec
provides a robust infrastructure for Oracle Real Application Clusters (RAC) that
simplifies management of RAC databases. SF Oracle RAC integrates existing
Veritas storage management and clustering technologies into a flexible solution
for administrators.
About SF Oracle RAC

SF Oracle RAC is a storage management and clustering solution that enables you
to:

■ Create a standard approach toward application and database management
in data centers. While other clusterware can only work with an Oracle
database, SF Oracle RAC incorporates existing Veritas storage management
and clustering technologies that provide flexible support for many types of
applications and databases. Administrators can apply existing expertise of
Veritas technologies toward this product.

■ Set up an infrastructure for Oracle RAC that simplifies database
management while fully integrating with Oracle Cluster Ready Services
(CRS).

■ Enhance scalability and availability with access to multiple RAC instances
per database in a cluster.

■ Transition from a local high availability site to a wide-area disaster recovery
environment with primary and secondary sites.

■ Back up and recover databases using volume-level and file system-level
snapshot technologies. SF Oracle RAC enables full volume-level snapshots
for off-host processing that reduce load on production systems, and file
system-level snapshots that involve efficient backup and rollback processes.

■ Prevent data corruption at the storage layer with robust split-brain   <<<<<<<<<<< This one is important
protection.

■ Increase scalability with high throughput and low latency technology for
use by Oracle Cache Fusion.

■ Share all types of files, in addition to Oracle database files, across nodes.

■ Increase availability and performance with dynamic multipathing (DMP),
which provides wide storage array support for protection from failures and
performance bottlenecks in the HBAs and SAN switches.

■ Model and test cluster configurations without affecting production systems
using the simulator and firedrill clustering technologies.

■ Optimize I/O performance through storage mapping technologies and
tunable attributes.


Gaurav