Return-Path: X-Original-To: oracle-l@orafaq.com Delivered-To: oracle-l@orafaq.com Received: from puck1183.startdedicated.com (localhost [127.0.0.1]) by puck1183.startdedicated.com (Postfix) with ESMTP id 986881960FB2 for ; Thu, 3 Mar 2016 20:19:22 +0100 (CET) Received: from turing.freelists.org (freelists-180.iquest.net [206.53.239.180]) by puck1183.startdedicated.com (Postfix) with ESMTPS for ; Thu, 3 Mar 2016 20:19:22 +0100 (CET) Received: from localhost (localhost [127.0.0.1]) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTP id BE3F43171E; Thu, 3 Mar 2016 14:19:20 -0500 (EST) X-Virus-Scanned: Debian amavisd-new at turing.freelists.org Received: from turing.freelists.org ([127.0.0.1]) by localhost (turing.freelists.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id bURPiZhcTBfl; Thu, 3 Mar 2016 14:19:20 -0500 (EST) Received: from turing.freelists.org (localhost [127.0.0.1]) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTP id 0CE95310B0; Thu, 3 Mar 2016 14:18:40 -0500 (EST) Received: with ECARTIS (v1.0.0; list oracle-l); Thu, 03 Mar 2016 14:17:58 -0500 (EST) Received: from localhost (localhost [127.0.0.1]) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTP id 69A1030A2D for ; Thu, 3 Mar 2016 14:17:58 -0500 (EST) Received: from turing.freelists.org ([127.0.0.1]) by localhost (turing.freelists.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id TmZKMcY+i9hE for ; Thu, 3 Mar 2016 14:17:58 -0500 (EST) Received: from nm50-vm8.bullet.mail.gq1.yahoo.com (nm50-vm8.bullet.mail.gq1.yahoo.com [67.195.87.248]) (using TLSv1 with cipher RC4-SHA (128/128 bits)) (No client certificate requested) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTPS id 1512A30AA9 for ; Thu, 3 Mar 2016 14:17:57 -0500 (EST) Received: from [127.0.0.1] by nm50.bullet.mail.gq1.yahoo.com with NNFMP; 03 Mar 2016 19:17:57 -0000 Received: from [216.39.60.180] by nm50.bullet.mail.gq1.yahoo.com with NNFMP; 03 Mar 2016 19:15:16 -0000 Received: from [98.139.170.178] by tm16.bullet.mail.gq1.yahoo.com with NNFMP; 03 Mar 2016 19:15:16 -0000 Received: from [98.139.212.215] by tm21.bullet.mail.bf1.yahoo.com with NNFMP; 03 Mar 2016 19:15:16 -0000 Received: from [127.0.0.1] by omp1024.mail.bf1.yahoo.com with NNFMP; 03 Mar 2016 19:15:16 -0000 Received: by 66.196.80.146; Thu, 03 Mar 2016 19:15:15 +0000 Date: Thu, 03 Mar 2016 19:15:15 +0000 From: Chris King Cc: Oracle-l Digest Users Message-ID: <1881571419.3145338.1457032515303.JavaMail.yahoo@mail.yahoo.com> In-Reply-To: <1AFD62082EEAF0448EF1815139687F134FDC282B@NC2PWEX504.us.ad.lfg.com> References: <1AFD62082EEAF0448EF1815139687F134FDC282B@NC2PWEX504.us.ad.lfg.com> Subject: Re: cost-effective, low-performance options for RAC? MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_3145337_701395341.1457032515280" X-archive-position: 63763 X-ecartis-version: Ecartis v1.0.0 Sender: oracle-l-bounce@freelists.org Errors-to: oracle-l-bounce@freelists.org X-original-sender: ckaj111@yahoo.ca Precedence: normal Reply-To: ckaj111@yahoo.ca List-help: List-unsubscribe: List-software: Ecartis version 1.0.0 List-Id: oracle-l X-List-ID: oracle-l List-subscribe: List-owner: List-post: List-archive: X-list: oracle-l X-Orig-Date: Thu, 3 Mar 2016 19:15:15 +0000 (UTC) ------=_Part_3145337_701395341.1457032515280 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Thanks everyone. This is an excellent discussion. In this particular case, I believe DataGuard is what this customer needs. B= ut they are afraid of possible data loss .I am aware of the maximum protect= ion mode, but if the network experiences the least bit of a glitch, failove= r will automatically occur. That seems more risky than setting parameters t= o write redo every 15 minutes, and then accept the possible 15 minutes of d= ata loss, should the primary fail moments before it is about to write the n= ext archived redo log. RAC doesn't protect against data loss as much as against the loss of one no= de. But compared side-by-side, if one server goes down, and it's a RAC node= , no data is lost. If the node that goes down is the primary server in a DG= configuration, there could be data loss. From: "Ruel, Chris" To: "tim@evdbt.com" ; "andrew.kerber@gmail.com" ; Jeremiah Wilton =20 Cc: "ckaj111@yahoo.ca" ; Oracle-l Digest Users Sent: Thursday, March 3, 2016 1:41 PM Subject: RE: cost-effective, low-performance options for RAC? =20 #yiv4832981647 #yiv4832981647 -- _filtered #yiv4832981647 {font-family:Helv= etica;panose-1:2 11 6 4 2 2 2 2 2 4;} _filtered #yiv4832981647 {panose-1:2 = 4 5 3 5 4 6 3 2 4;} _filtered #yiv4832981647 {font-family:Calibri;panose-1:= 2 15 5 2 2 2 4 3 2 4;}#yiv4832981647 #yiv4832981647 p.yiv4832981647MsoNorma= l, #yiv4832981647 li.yiv4832981647MsoNormal, #yiv4832981647 div.yiv48329816= 47MsoNormal {margin:0in;margin-bottom:.0001pt;font-size:12.0pt;color:black;= }#yiv4832981647 a:link, #yiv4832981647 span.yiv4832981647MsoHyperlink {colo= r:blue;text-decoration:underline;}#yiv4832981647 a:visited, #yiv4832981647 = span.yiv4832981647MsoHyperlinkFollowed {color:purple;text-decoration:underl= ine;}#yiv4832981647 span.yiv4832981647EmailStyle17 {color:#1F497D;}#yiv4832= 981647 .yiv4832981647MsoChpDefault {font-size:10.0pt;} _filtered #yiv483298= 1647 {margin:1.0in 1.0in 1.0in 1.0in;}#yiv4832981647 div.yiv4832981647WordS= ection1 {}#yiv4832981647 Maybe it=E2=80=99s just a matter of personal diffe= rences with definitions of HA, but, I have always professed to customers th= at HA is one of the hallmark features of RAC (along with scale out) as it a= llows properly designed applications to take advantage of failover for both= planned and unplanned outages through TAF or a mechanisms like Active Grid= link for Oracle through Weblogic.=C2=A0 Things like rolling patches or hard= ware failures at the node level can be sustained without outage via a prope= r RAC implementation. =C2=A0 On the other hand, to me, Data Guard is not a= s much an HA solution as it is a DR solution.=C2=A0 Data Guard, no matter h= ow hard you try, will not provide seamless failover.=C2=A0 I will admit, th= at what I am saying is very academic as it is rare that I have encountered = applications that have been properly designed for take advantage of RAC=E2= =80=99s TAF for failed transactions.=C2=A0 Seems like most apps these days = use JDBC thin clients that automatically preclude you from using TAF unless= you have them connected through a Weblogic connection pool that is protect= ed by Active Gridlink for Oracle. =C2=A0 So, even a customer, as described= below, could take advantage of/need RAC if that have a need for Five 9=E2= =80=99s and their application is appropriately designed.=C2=A0 Which brings= up another argument, half the time, people that think they need five 9=E2= =80=99s don=E2=80=99t, or, only one component of their application infrastr= ucture has it (i.e. the database, but, not the storage or app-tier). =C2= =A0 My 2 cents anyway. =C2=A0 Chris.. =C2=A0 ____________________________= _________________________________________ Chris Ruel * Oracle Database Admi= nistrator * Lincoln Financial Group cruel@lfg.com * Desk:317.759.2172=C2=A0= * Cell 317.523.8482 =C2=A0 From: oracle-l-bounce@freelists.org [mailto:ora= cle-l-bounce@freelists.org]On Behalf Of Tim Gorman Sent: Thursday, March 03, 2016 1:17 PM To: andrew.kerber@gmail.com; Jeremiah Wilton Cc: ckaj111@yahoo.ca; Oracle-l Digest Users Subject: Re: cost-effective, low-performance options for RAC? =C2=A0 HA is= not a practical use-case (by itself) for RAC.=C2=A0 If HA is their require= ment, they should use Data Guard. Performance scalability is the practical use-case for RAC.=C2=A0 Your custo= mer has self-selected themselves out of that use-case, but apparently they'= d rather spend more money, time, and effort to get to the same place.=C2=A0= C'est la vie... On 3/3/16 10:46, Andrew Kerber wrote:=20 You can run RAC over iscsi and dnfs according to the oracle documentation.= =C2=A0 But I agree with you, sounds like they probably dont need it that ba= dly, unless its solely for HA. =C2=A0 On Thu, Mar 3, 2016 at 11:37 AM, Jer= emiah Cetlin Wilton wrote:=20 You could build something like this on smaller EC2 instances. =C2=A0 http:= //aws.amazon.com/articles/7455908317389540 =C2=A0 Jeremiah From:"Chris Kin= g" To: "Oracle-l Digest Users" Sent: Thursday, March 3, 2016 8:54:11 AM Subject: cost-effective, low-performance options for RAC? I was asked an od= d question today. A customer is determined to use an Oracle RAC solution, a= nd we are trying to convince them that in their case, this is too big a sol= ution for what they need. In the meantime, however, I've been asked a few i= nteresting questions, which I would love to have your feedback on. =C2=A0 = The customer's requirement is for a small database, with only 200 users max= imum, and not concurrent users. We're not looking for high-performance beca= use the database will not be heavily used. =C2=A0 I've always used fibre = connects with Oracle RAC. Is it possible to run RAC with something other th= an fibre? I've been asked if we could run RAC with a single SAS attached di= sk array.. I'm assuming not, but maybe there are other lower-cost RAC compa= tible storage and network connectivity options. =C2=A0 May I have your fee= dback on this? =C2=A0 Thanks! -ChrisK =C2=A0 =C2=A0 =C2=A0=20 -- Andrew W. Kerber 'If at first you dont succeed, dont take up skydiving.'=20 =C2=A0 Notice of Confidentiality: **This E-mail and any of its attachments= may contain=20 Lincoln National Corporation proprietary information, which is privileged, = confidential, or subject to copyright belonging to the Lincoln National Corporation famil= y of=20 companies. This E-mail is intended solely for the use of the individual or = entity to=20 which it is addressed. If you are not the intended recipient of this E-mail= , you are=20 hereby notified that any dissemination, distribution, copying, or action ta= ken in=20 relation to the contents of and attachments to this E-mail is strictly proh= ibited=20 and may be unlawful. If you have received this E-mail in error, please noti= fy the=20 sender immediately and permanently delete the original and any copy of this= E-mail=20 and any printout. Thank You.** ------=_Part_3145337_701395341.1457032515280 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Thanks everyone. Thi= s is an excellent discussion.

In this part= icular case, I believe DataGuard is what this customer needs. But they are = afraid of possible data loss .I am aware of the maximum protection mode, bu= t if the network experiences the least bit of a glitch, failover will autom= atically occur. That seems more risky than setting parameters to write redo= every 15 minutes, and then accept the possible 15 minutes of data loss, sh= ould the primary fail moments before it is about to write the next archived= redo log.

RAC doesn't protect again= st data loss as much as against the loss of one node. But compared side-by-= side, if one server goes down, and it's a RAC node, no data is lost. If the= node that goes down is the primary server in a DG configuration, there cou= ld be data loss.



From: "Ruel, Chris" <Chris.Ruel@lf= g.com>
To: "tim@evd= bt.com" <tim@evdbt.com>; "andrew.kerber@gmail.com" <andrew.kerber@= gmail.com>; Jeremiah Wilton <jcwilton93@earlham.edu>
Cc: "ckaj111@yahoo.ca" <ckaj111@= yahoo.ca>; Oracle-l Digest Users <oracle-l@freelists.org>
<= span style=3D"font-weight: bold;">Sent: Thursday, March 3, 2016 = 1:41 PM
Subject: RE: c= ost-effective, low-performance options for RAC?

Maybe it=E2=80=99s just a matter of personal differences with definitions= of HA, but, I have always professed to customers that HA is one of the hal= lmark features of RAC (along with scale out) as it allows properly designed ap= plications to take advantage of failover for both planned and unplanned out= ages through TAF or a mechanisms like Active Gridlink for Oracle through We= blogic.  Things like rolling patches or hardware failures at the node level can be sustained without outage via= a proper RAC implementation.
=20
 
=20
On = the other hand, to me, Data Guard is not as much an HA solution as it is a = DR solution.  Data Guard, no matter how hard you try, will not provide= seamless failover.  I will admit, that what I am saying is very academic as it= is rare that I have encountered applications that have been properly desig= ned for take advantage of RAC=E2=80=99s TAF for failed transactions.  = Seems like most apps these days use JDBC thin clients that automatically preclude you from using TAF unless you have them connec= ted through a Weblogic connection pool that is protected by Active Gridlink= for Oracle.
=20
&n= bsp;
=20
So,= even a customer, as described below, could take advantage of/need RAC if t= hat have a need for Five 9=E2=80=99s and their application is appropriately= designed.  Which brings up another argument, half the time, people that think they need fiv= e 9=E2=80=99s don=E2=80=99t, or, only one component of their application in= frastructure has it (i.e. the database, but, not the storage or app-tier).<= /span>
=20
&n= bsp;
=20
My = 2 cents anyway.
=20
&n= bsp;
=20
Chr= is..
=20
&n= bsp;
=20
___= __________________________________________________________________=20
Chr= is Ruel * Oracle Database Administrator * Lincoln Financial Group=20
cruel@lfg= .com * Desk:317.759.217= 2 * Cell 317.523.8482
=20
&n= bsp;
=20
= From: oracle-l-bounce@freelist= s.org [mailto:oracle-l-bounce@freelists.org] On Behalf Of Tim Gorman
Sent: Thursday, March 03, 2016 1:17 PM
To: andrew.kerber@gmail.com; Jeremiah Wilton
Cc: ckaj111@yahoo.ca; Oracle-l Digest Users
Subject: Re: cost-effective, low-performance options for RAC?=
=20
 
=20
HA is= not a practical use-case (by itself) for RAC.  If HA is their require= ment, they should use Data Guard.

Performance scalability is the practical use-case for RAC.  Your custo= mer has self-selected themselves out of that use-case, but apparently they'= d rather spend more money, time, and effort to get to the same place. = C'est la vie...



=20
On 3/3/16 10:46, Andrew Kerber wrote:=
=20
You can run RAC over iscsi and dnfs a= ccording to the oracle documentation.  But I agree with you, sounds li= ke they probably dont need it that badly, unless its solely for HA.
= =20
 
=20
On Thu, Mar 3, 2016 at 11:37 AM, Jere= miah Cetlin Wilton <jcwilton93@earlham.edu> wrote:
=20
You could build something like this o= n smaller EC2 instances.
=20
 
=20
 
=20
Jerem= iah
=20
From: "Chris King" <ckaj111@yahoo.ca>
To: "Oracle-l Digest Users" <oracle-l@freelists.org>
Sent: Thursday, March 3, 2016 8:54:11 AM
Subject: cost-effective, low-performance options for RAC?=20
I was asked an odd question today. A customer is d= etermined to use an Oracle RAC solution, and we are trying to convince them= that in their case, this is too big a solution for what they need. In the meantime, however, I= 've been asked a few interesting questions, which I would love to have your= feedback on.
=20
 
=20
The customer's requirement is for a small database= , with only 200 users maximum, and not concurrent users. We're not looking = for high-performance because the database will not be heavily used.
=20
 
=20
I've always used fibre connects with Oracle RAC. I= s it possible to run RAC with something other than fibre? I've been asked i= f we could run RAC with a single SAS attached disk array.. I'm assuming not, but maybe there are o= ther lower-cost RAC compatible storage and network connectivity options.
=20
 
=20
May I have your feedback on this?
=20
 
=20
Thanks!
=20
-ChrisK
=20
 
=20
 
=20
 
=20



--
=20
Andrew W. Kerber

'If at first you dont succeed, dont take up skydiving.'
=20
 
=20
Notice of Confidentiality: **This E-mail and any of its attachments ma= y contain
Lincoln National Corporation proprietary infor= mation, which is privileged, confidential,
or subject to = copyright belonging to the Lincoln National Corporation family of
companies. This E-mail is intended solely for the use of the indi= vidual or entity to
which it is addressed. If you are no= t the intended recipient of this E-mail, you are
hereby = notified that any dissemination, distribution, copying, or action taken in =
relation to the contents of and attachments to this E-ma= il is strictly prohibited
and may be unlawful. If you ha= ve received this E-mail in error, please notify the
send= er immediately and permanently delete the original and any copy of this E-m= ail
and any printout. Thank You.**

=
------=_Part_3145337_701395341.1457032515280-- -- http://www.freelists.org/webpage/oracle-l