most of our unplanned outages are related to either network or SAN related
(ie HBA's not correctly virtualizing the LUNS from the SAN). RAC would
have helped with some of these.
most of our planned outages are related to OS upgrades & patches, Planned
network segment outages, SAN Switch maintenance etc. RAC would also help
with most of these. We will often failover our critical oracle apps using
standard physical standby.
the other major planned outage cause is oracle upgrades/patches - RAC
would NOT help with these (maybe in the future???)
are many folks out there actually using RAC in a demanding prod
environment?
steve
Niall Litchfield <niall.litchfield@xxxxxxxxx>
09/13/2004 02:35 PM
Please respond to Niall Litchfield
To: Stephen Evans <evans036@xxxxxxxxxxx>
cc: oracle-l@xxxxxxxxxxxxx
Subject: Re: Is RAC really HA on Linux
Well I received a quote off-list suggesting that there was a point of view
that
"Instance Failures account for a huge majority of
the outages"
If you share this viewpoint, which I don't, then RAC/OPS might count
as an HA solution.
--
Niall Litchfield
Oracle DBA
http://www.niall.litchfield.dial.pipex.com
--
http://www.freelists.org/webpage/oracle-l