Re: A DBA philosopical question

From: HansF <News.Hans_at_telus.net>
Date: Sat, 03 Dec 2005 01:38:05 GMT
Message-Id: <pan.2005.12.03.01.38.04.725995@telus.net>

On Sat, 03 Dec 2005 01:18:46 +0000, Bob Jones wrote:

>
> "HansF" <News.Hans_at_telus.net> wrote in message
> news:pan.2005.12.02.23.25.39.864182_at_telus.net...

>> On Fri, 02 Dec 2005 22:48:59 +0000, Randy Harris wrote:
>>
>>>
>>> Thanks, an excellent article. Mr. Kagel doesn't seem very fond of RAID 5.
>>
>> He's not the only one - look at the list of BAARFers
>>

>
> I have read many articles like that. They always seem to lack practicality.
> The notion of "RAID 5 is bad" is very misleading. It has its pros and cons
> just like any other RAIDs. Whether to use it is heavily application
> dependent.
>

It's RAID, therefore it must be good, right? After all, it's buzzword compliant and there are papers of various colours describing it.

And it's the cheapest of the RAIDs that purports to give some protections, so that makes it excellent!

The above is as far as any non-technical manager wants to take the discussion - and hopefully that satisfies the balance between techno-dweebs and accounting-trolls.

As you say, it's application dependent. The problem is, unless there is heavy use of memory in front of the disk, a database application is not among the list of optimal uses.

> If RAID 5 is all that bad, why is it so widely used and supported?

MacDonalds is widely used and heavily supported, as is WalMart. Are they good as well? Does the populist model define good, or 'popular therefore with minimum social risk'?

-- 
Hans Forbrich                           
Canada-wide Oracle training and consulting
mailto: Fuzzy.GreyBeard_at_gmail.com   
*** Top posting [replies] guarantees I won't respond. ***

Received on Fri Dec 02 2005 - 19:38:05 CST