Thanks for the kind words John.
The direct link is at
http://www.tech15.ukoug.org/default.asp?p=12861&dlgact=shwprs&prs_prsid=11011&day_dayid=95
. You'll need to be a UKOUG member to download. There's also an excellent
presentation (which I reference) available at
http://www.doag.org/formes/servlet/DocNavi?action=getFile&did=6056763&key=
from Robert Bialek of Trivadis.
A lot of the operational concerns come down to the fact that you need to
use your clusterware commands and scripts to control the database
availability rather than the traditional Oracle tools. This typically means
changing operational procedures somewhat - as an example consider a
failover to a standby site. You'll need to ensure you can do this without
the clusterware restarting the old primary halfway through the operation
and so on.
If you are considering licensing the site as a cold failover site, you'll
need to be aware of exactly what the 10 day rule means, and you will need
to keep track of this. It's really surprisingly easy to run unexpectedly on
the unlicensed node without anyone noticing..
On Sat, Jan 30, 2016 at 6:09 PM, Nagaraj S <nagaraj.chk@xxxxxxxxx> wrote:
Thank You John Thomas, Do you have link or ppt of Niall
Litchfield presentation
On Sat, Jan 30, 2016 at 11:29 PM, John Thomas <jt2354@xxxxxxxxx> wrote:
Niall Litchfield did a very good presentation at last year's UKOUG on
using Grid/Clusterware for failover to a cold instance operating under the
ten day rule.
Regards
JT
On Sat, 30 Jan 2016, 17:53 Nagaraj S <nagaraj.chk@xxxxxxxxx> wrote:
Good Morning Gurus,
I need help on RAC Vs Non-RAC Failover Design . One of our application
need to move out of RAC due to cost factor in this case which HA will be
better? I am thinking of OS Failover with Active Passive mechanism. Also
help me on Application perspective what are the things I need to check
before moving it to NON RAC?
-nagaraj
