RE: Is RAC DOA?

  • From: "Mark W. Farnham" <mwf@xxxxxxxx>
  • To: <oracle-l@xxxxxxxxxxxxx>
  • Date: Mon, 16 Aug 2004 11:48:10 -0400

One slight quibble, regarding increased availability.

The maximum theoretical increased availability is the time a single node
would have been down unexpectedly, plus the time of outage to execute a
failover to a stand by node for preventive maintenance, MINUS the time of
outage due to increased complexity (cluster failures, multi-node disk
sharing, single node failure that brings down the cluster anyway, etc.,
etc., etc., go Yul Brynner!)

I don't have any cases where this number does not come out negative (yet),
largely due to the frequency of cluster failures and cluster functionality
destroying events such as multi-node listener confusion. I think a fair
amount of the down time is due to a combination of pilot error and pilot
misinformation (documentation errors and omissions.) I count pilot error on
the MINUS side of the equation when it is pilot error that could not have
happened on a single node implementation. My hopeful expectation is that
grids will become nearly as bulletproof as single nodes in the not too
distant future, especially when given adequate guidance from expensive
billed by hour resources.

Now V6.2 on a stable vax cluster is a different issue, but since the DLM
threaded ploddingly through the disk farm, that is an entirely different
issue.

mwf
-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx]On Behalf Of Freeman, Donald
Sent: Monday, August 16, 2004 11:29 AM
To: oracle-l@xxxxxxxxxxxxx
Subject: RE: Is RAC DOA?


One thing to do is check the archives for this forum.  There have been a =
lot of discussions about RAC.  I'm a bit glum about it.  I'm pretty new =
as an Oracle DBA and we have already installed RAC on our transactional =
system and have allocated money for our datawarehouse. At the time I =
didn't know enough to influence those decisions. I have observed that =
the transactional dba's have had some trouble keeping both nodes up when =
they were supposed to be up.

What I have extracted out of the conversations here is that your =
management problems will increase, your cost will increase, your =
performance will decrease.   You should gain some increased  =
availability.  If I didn't summarize that right somebody should correct =
me.=20

-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx]On Behalf Of Ray Stell
Sent: Monday, August 16, 2004 11:17 AM
To: oracle-l@xxxxxxxxxxxxx
Subject: Is RAC DOA?



I posed a question here last week wrt RAC and got one response.
Is there a concentration of RAC users elsewhere?  I find the=20
metalink forum practically worthless, which is probably another
strong indicator that RAC is dead.  Pricing issues, right?=20
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
Ray Stell   stellr@xxxxxx     (540) 231-4109     KE4TJC    28^D
----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to:  oracle-l-request@xxxxxxxxxxxxx
put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------
----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to:  oracle-l-request@xxxxxxxxxxxxx
put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------


----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to:  oracle-l-request@xxxxxxxxxxxxx
put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------

Other related posts: