Measures and Rules

Question

Measures and Rules

Grant Fritchey

SSC Guru

Points: 398690
More actions
February 6, 2010 at 11:50 am

#141089

Comments posted to this topic are about the item Measures and Rules
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning

Viewing 9 posts - 1 through 8 (of 8 total)

You must be logged in to reply to this topic. Login to reply

Tobar SSCarpal Tunnel Points: 4876 More actions · Answer 1

Great Article.

Each person must assess their own needs and responses when considering "cookie cutter" numbers. As much as we would like to believe, we are not a one size fits all creature. But we like to think that way. The human body is infinitely more complex than a database, but the same rule applies. Too often after implementing a cookie cutter number, people/DBAs only question the number if things go badly, and there is nothing wrong with that, but that doesn't mean it is the wrong number for everybody else. We hope that the "experts" know what they are talking about, and mostly they do. That is not to say that through research you can not come up with a better number for the situations you are in.

<><
Livin' down on the cube farm. Left, left, then a right.

jay-h SSCoach Points: 18816 More actions · Answer 2

It's interesting (and common) that the diatribe against the original article is turned into a gender flame war.

The initial point of Paulos argument (and indeed some of his excellent books) is how misleading our gut instincts can be about statistics (leave it to that commenter to decide that he is derogoagory toward women).

This same kind of discussion has occurred in other areas of medical screening too (hint: similar advice was given for other non gender related diseases as well), and in our "precautionary prinicple" driven society, the gut reaction is often wrong, and sometimes counter productive.

When it comes to understanding the nuances of statistics, should I listen to a top mathematician in the field or a law professor... hmmm which to choose?

...

-- FORTRAN manual for Xerox Computers --

Grant Fritchey SSC Guru Points: 398690 More actions · Answer 3

Tobar (2/8/2010)
Great Article.
Each person must assess their own needs and responses when considering "cookie cutter" numbers. As much as we would like to believe, we are not a one size fits all creature. But we like to think that way. The human body is infinitely more complex than a database, but the same rule applies. Too often after implementing a cookie cutter number, people/DBAs only question the number if things go badly, and there is nothing wrong with that, but that doesn't mean it is the wrong number for everybody else. We hope that the "experts" know what they are talking about, and mostly they do. That is not to say that through research you can not come up with a better number for the situations you are in.

Thanks, and I agree with you. If you really dig, probably, each and every measurement is unique to a given situation, but the variation may be very slight, so does it matter that your disk queue length is 4.5 or 5 before you start get a little concerned? But sometimes, those measurements can vary wildly from one situationn to another. The more you know, the better you'll be able to understand if the measurement applies in your situation. The problem is, there's a lot to know.

"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt

Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning

Andy Leonard SSChampion Points: 10134 More actions · Answer 4

Great editorial Grant!

This reminded me of the demise of my checklist. I used to think database tuning was a matter of checking off items on my checklist, in order, as I made corrections. I believed this until some of my changes took a server to its knees.

Investigation revealed the checklist wasn't the problem - my faith in it was the problem.

I'd built the checklist from experience. I went back to that experience and instead documented the methodology I'd used to gather measurements and create tests. This methodology replaced my checklist.

:{> Andy

Andy Leonard, Chief Data Engineer, Enterprise Data & Analytics

Grant Fritchey SSC Guru Points: 398690 More actions · Answer 5

Andy Leonard (2/8/2010)
Great editorial Grant!
This reminded me of the demise of my checklist. I used to think database tuning was a matter of checking off items on my checklist, in order, as I made corrections. I believed this until some of my changes took a server to its knees.
Investigation revealed the checklist wasn't the problem - my faith in it was the problem.
I'd built the checklist from experience. I went back to that experience and instead documented the methodology I'd used to gather measurements and create tests. This methodology replaced my checklist.
:{> Andy

Now that sounds like a great presentation. When are you giving it?

"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt

Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning

Andy Leonard SSChampion Points: 10134 More actions · Answer 6

Grant Fritchey (2/8/2010)
Now that sounds like a great presentation. When are you giving it?

I love you man!

:{>

Andy Leonard, Chief Data Engineer, Enterprise Data & Analytics

GSquared SSC Guru Points: 260824 More actions · Answer 7

Standard measures and rules of that sort are of incredible value (in most cases) to people who are first learning a subject. Once one really knows the subject and understands it, one can violate the "rules", because one understands enough to do so in a positive way.

- Gus "GSquared", RSVP, OODA, MAP, NMVP, FAQ, SAT, SQL, DNA, RNA, UOI, IOU, AM, PM, AD, BC, BCE, USA, UN, CF, ROFL, LOL, ETC
Property of The Thread

"Nobody knows the age of the human race, but everyone agrees it's old enough to know better." - Anon

SQLRNNR SSC Guru Points: 281334 More actions · Answer 8

Great Editorial. You bring up a very simple point that we try to make daily. There are measures out there that are guides. The guide is a starting point and you must test to find what works best in your environment. That too may change over time - so as Andy said - it is really the methodology that one needs to implement to make sure the measures and guides established for their environment are correct.

Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events