Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: tough choices

Re: tough choices

From: Howard J. Rogers <hjr_at_dizwell.com>
Date: Sat, 19 Jun 2004 07:55:01 +1000
Message-ID: <40d3649a$0$18671$afc38c87@news.optusnet.com.au>

"Michael Austin" <maustin_at_firstdbasource.com> wrote in message news:gOGAc.6827$1u1.2440_at_newssvr24.news.prodigy.com...

> >>According to Oracle marketing and technical folks (2 years ago), this
> >>really only worked as advertised on 2 platforms. Can you guess which
> >>ones they were??
> >
> >
> > You'll have to refresh my memory, but from my own personal experience,
two
> > years ago, TAF was working as advertised on Windows and Solaris and
Linux.
> > And Tru64, too.
>
> Tru64 and OpenVMS... the others had to do some gyrations with the
> filesystems on Windows and Solaris after the original node crashed...
> IIRC -- didn't take long, just longer than the reuqired instance
> recovery on a surviving node. As I said this was 2 years ago and the
> grey-matter gets clouded sometimes... :)

Fair enough. But let's be clear: regardless of your operating system or file system, resources in use at the time of a node/instance failure are always locked until instance recovery has completed on other nodes (performed automatically by LMON, incidentally). So, no matter your operating system, or the file system, the resumption of a select statement will always "take longer than the instance recovery" if what is being selected is not 100% clean (ie, if it has also been subjected to DML, thus necessitating recovery). A clean buffer needs no recovery, and so the block on disk it represents can be made available to surviving nodes more quickly than a dirty one (remastering may still be needed, even so).

There are no more "gyrations" involved with a file system on Windows and Solaris than there is with any file system-based RAC installation on any other O/S.

The only thing I can think of that Oracle "marketing and technical folks" might have been saying 2 years ago was that you had to RAC with raw devices unless you had Tru64 or VMS, because no cluster file systems were widely available for other platforms. One of Tru64's great claims to fame at that time was. indeed, that it was the only Unix for which a cluster file system was available. That rapidly became "no longer the case". And, of course, Oracle itself wrote a cluster file system for Windows and Linux.

TAF, nevertheless, has nothing whatsoever to do with the way the database is physically stored, and therefore the question of whether you are RAC'ing with raw or a cluster file system has no bearing on its capabilites... and never did.

Regards
HJR Received on Fri Jun 18 2004 - 16:55:01 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US