RE: RAC in NAS

How much effort did you put into designing your application for RAC? Do you put any effort to reduce data block passing back and forth between nodes?

Are there any benchmarks showing performance of RAC systems: 1. Data blocks isolated to individual nodes so as to limit passing blocks and forth 2. Little to no effort to isolate blocks to nodes

and compare how they scale? I have not see any. Just wondering if there are any out there.

• Original message -------------- From: "Matthew Zito" <mzito_at_gridapp.com>

Not to start a holy war on this, but it *IS* possible to have very successful horizontal scaling using RAC. One of our appliance customers has scaled to a 10-node cluster linearly, saving them over $800,000 just in infrastructure costs, even counting the additional cost of the RAC licenses and our software. We have several other customers that have scaled to 4-node clusters easily and saved *only* $100k-200k.

For sure, RAC is not a magical solution, but there are a number of good reasons to consider it, such as HA, reduced infrastructure cost, and incremental scalability. There's increased complexity surrounding it as well, which is why there's folks like my company and Kevin Closson's, Polyserve, whose value proposition is reducing that complexity.

Thanks,
Matt

--
Matthew Zito
Chief Scientist
GridApp Systems
P: 646-452-4090
mzito_at_gridapp.com
http://www.gridapp.com



-----Original Message-----
From: oracle-l-bounce_at_freelists.org on behalf of Mladen Gogala
Sent: Wed 7/26/2006 9:29 PM
To: pythianbrinsmead_at_gmail.com
Cc: Yavor_Ivanov_at_stemo.bg; oracle-l_at_freelists.org
Subject: Re: RAC in NAS


On 07/26/2006 09:10:03 PM, Mark Brinsmead wrote:


>

> A more relevant question to ask is whether the customer actually *needs*

> RAC.  Or 8 CPUs.  If they had not done enough research to know whether the

> storage layer they had chosen would even work, then chances are, they have

> not sized the hardware nor adequately researched the requirements for RAC,

> either.


Horizontal scaling is also known as crucifixion. RAC implementation without
careful planning and analysis is just that.





>

> If you're *certain* that you *must* use NAS (NFS), then be certain you do

> your homework when choosing your other components.  (Oops.  That's how we

> got here, isn't it?)  Redhat cannot (yet) support Asynch I/O with NFS -- the

> same may be true of other supported Linux distros, but I'm not certain.

> Perhaps you could use Solaris?  [Ooh!  Wouldn't *that* annoy IBM... ;-) ]


Asynch I/O is not that important. Oracle can emulate it using I/O slaves. Granted,
it's not as good as the real thing, but you will not sufer much, either. Direct I/O
is much more important and it is supported. FC5 is the sign of things to come. It
does support full NFS4 version, with cient caching and async I/O included. EL5 is
likely to have those features, minus bugs, discovered by free beta testers, like
me.



--
Mladen Gogala
http://www.mgogala.com

--
http://www.freelists.org/webpage/oracle-l
--
http://www.freelists.org/webpage/oracle-l