RE: Oracle 12c and All-Flash Technology

From: Mark W. Farnham <mwf_at_rsiz.com>
Date: Sat, 13 Feb 2016 19:12:09 -0500
Message-ID: <01a601d166bc$58b9c730$0a2d5590$_at_rsiz.com>

Yet the specific question was about testing the media stack.

If you want to know whether your whole application needs that throughput, then I'd agree with you. But if you're testing the capacity of a component of performance for determine price performance, you may not even have a specific application suite in mind yet.

mwf

-----Original Message-----
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Mladen Gogala Sent: Friday, February 12, 2016 8:06 PM
To: oracle-l_at_freelists.org
Subject: Re: Oracle 12c and All-Flash Technology

On 02/12/2016 10:20 AM, Mark W. Farnham wrote:
> And specifically for testing the throughput with respect to use by Oracle of a persistent media stack, there is SLOB from Kevin Closson.
>
> It's Silly. It's Little. It just drives Oracle i/o against pretty much any media stack. It's free and easy to use.
>
> Be a SLOBer. They've even got a cool button that lights up you can wear.
>
> (Not a paid endorser. Just a friend and fan of Kevin.)
>
> mwf
>

With all due respect to you and Kevin, I disagree with that advice. I would always recommend testing using a test that resembles the expected workload, using home grown applications and tools like swingbench or hammerora. SLOB simulates Oracle use really well, but nothing simulates the real thing like the real thing.

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


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

Received on Sun Feb 14 2016 - 01:12:09 CET