April 27, 2011 at 11:13 pm
Comments posted to this topic are about the item Resource Governor Calculations
April 28, 2011 at 6:46 am
I had to look at this a few times to make sure I wasn't missing something. I kept expecting the set max to need to be modified by something in this situation to make the question a little trickier.
April 28, 2011 at 7:05 am
I thought I would get this wrong but managed to score. If 90 was given as an answer option, I would have selected it.
M&M
April 28, 2011 at 7:27 am
When attempting to learn the effective max% concept I read the following SQLServer Central article which I find misleading (at least as I interpreted it): http://www.sqlservercentral.com/articles/Resource+Governor/64034/
It provides the following definition: "... the Effective Maximum Rate. It is equal to the difference between the maximum value of said resource pool and sum of minimum values of all others."
Using this definition the Sales pool would calculate as follows: min(75,(75-10)) or min(75,65) which is 65. As this was not one of the options available as an answer I knew something was amiss in my understanding of the topic. As I read your referenced article on MSDN, it aligns with the explanation given in the answer and makes sense to me.
Am I reading the SQL Server Central article wrong? Either way, thanks for the education.
Chris Umbaugh
Data Warehouse / Business Intelligence Consultant
twitter @ToledoSQL
April 28, 2011 at 7:44 am
Chris Umbaugh (4/28/2011)
When attempting to learn the effective max% concept I read the following SQLServer Central article which I find misleading (at least as I interpreted it): http://www.sqlservercentral.com/articles/Resource+Governor/64034/It provides the following definition: "... the Effective Maximum Rate. It is equal to the difference between the maximum value of said resource pool and sum of minimum values of all others."
Using this definition the Sales pool would calculate as follows: min(75,(75-10)) or min(75,65) which is 65. As this was not one of the options available as an answer I knew something was amiss in my understanding of the topic. As I read your referenced article on MSDN, it aligns with the explanation given in the answer and makes sense to me.
Am I reading the SQL Server Central article wrong? Either way, thanks for the education.
I read both that article and the one Steve referenced. I made the same mistake in reading of the first one. However the second one that Steve references and has charts similar to his states this....
"If a pool has a nonzero MIN defined the effective MAX value of other pools is readjusted as the minimum of the configured MAX value of a pool and the sum of the MIN values of other pools subtracted from 100 percent."
So aparently you do not subtract the other values from the Max of the pool, you subtract from 100. Makes sense becuase how can you have a max more than the leftover resources or 100% - (total of other pool Min setting).
Nice question Steve. Looks like almost 30% are "learning" today!
April 28, 2011 at 7:59 am
Definitely had to read that explanation a few times. Thanks.
April 28, 2011 at 8:35 am
It's a fairly arbitrary concept and with conflicting information on the calculations, it's hard to understand what is correct. The way that it's given in the white paper, which is where a few people pointed me, is what seems to be the most accepted way. I suspect the SSC article is misinterpreting it a bit and I'll look at correcting that.
Glad you liked the question.
April 28, 2011 at 9:58 am
Good question and I learned something.
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
April 28, 2011 at 10:23 am
Chris Umbaugh (4/28/2011)
When attempting to learn the effective max% concept I read the following SQLServer Central article which I find misleading (at least as I interpreted it): http://www.sqlservercentral.com/articles/Resource+Governor/64034/It provides the following definition: "... the Effective Maximum Rate. It is equal to the difference between the maximum value of said resource pool and sum of minimum values of all others."
Using this definition the Sales pool would calculate as follows: min(75,(75-10)) or min(75,65) which is 65. As this was not one of the options available as an answer I knew something was amiss in my understanding of the topic. As I read your referenced article on MSDN, it aligns with the explanation given in the answer and makes sense to me.
Am I reading the SQL Server Central article wrong? Either way, thanks for the education.
All pools are guaranteed the minimum. So I calculate it as 100%-(total of all minimums EXCEPT for the pool I am calculating for, in this case we only have a total of 10 assigned to OTHER pools and we do not use the 25% minimum since we are looking at MAX for the Sales pool) which equals 90% OR the assigned value 75%, whichever is smaller.
Short version: 100-(10(admin)+0(default))=90 OR 75, whichever is smaller.
Or shorter: min(75,(100-(10))
Great question!!
Peter Trast
Microsoft Certified ...(insert many literal strings here)
Microsoft Design Architect with Alexander Open Systems
April 28, 2011 at 10:37 am
Steve Jones - SSC Editor (4/28/2011)
It's a fairly arbitrary concept and with conflicting information on the calculations, it's hard to understand what is correct. The way that it's given in the white paper, which is where a few people pointed me, is what seems to be the most accepted way. I suspect the SSC article is misinterpreting it a bit and I'll look at correcting that.Glad you liked the question.
I am fairly sure that some of the calculations in the article http://www.sqlservercentral.com/articles/Resource+Governor/64034/ are incorrect. When calculating the Effective MAX for a pool you do NOT include the MIN for the pool you are calculating on, but you subtract the total minimums of all other pools from 100%, NOT the assigned MAX for the pool. This is one of the biggest sources of confusion on the Resource Governor. See my other post on this thread for more...
Unless the Microsoft Official Curriculum is wrong... that could never happen...
Peter Trast
Microsoft Certified ...(insert many literal strings here)
Microsoft Design Architect with Alexander Open Systems
April 28, 2011 at 10:43 am
Peter Trast (4/28/2011)
I am fairly sure that some of the calculations in the article http://www.sqlservercentral.com/articles/Resource+Governor/64034/ are incorrect. When calculating the Effective MAX for a pool you do NOT include the MIN for the pool you are calculating on, but you subtract the total minimums of all other pools from 100%, NOT the assigned MAX for the pool. This is one of the biggest sources of confusion on the Resource Governor. See my other post on this thread for more...Unless the Microsoft Official Curriculum is wrong... that could never happen...
LOL Peter... Nice one.
Seriously though there seems to be enough confusion about this calculation to cause more than one person to mis quote it.
Thanks for confirming what is in the current Microsoft Official Curriculum
April 28, 2011 at 11:39 pm
Great question, but actually an easy one to get correct. If you just read the question and see "effective Max CPU %" and you look at the table and see 75%, it is easily to think:
"Duh, of course it is 75%, it is right there in the table!". And it is the right answer 🙂
A more tricky question would have been if the answers included 90, as stated before, or if the minimums of the other pools had been higher.
Nonetheless, still a great question. Thanks.
Need an answer? No, you need a question
My blog at https://sqlkover.com.
MCSE Business Intelligence - Microsoft Data Platform MVP
April 29, 2011 at 5:47 am
Koen Verbeeck (4/28/2011)
Great question, but actually an easy one to get correct. If you just read the question and see "effective Max CPU %" and you look at the table and see 75%, it is easily to think:"Duh, of course it is 75%, it is right there in the table!". And it is the right answer 🙂
A more tricky question would have been if the answers included 90, as stated before, or if the minimums of the other pools had been higher.
Nonetheless, still a great question. Thanks.
It's funny becasue that was exactly my first reaction. "Duh, it's 75!", and I would have gotten the 2 points but I would not have been correct and I would not have learned anything. It was a good question.
Chris Umbaugh
Data Warehouse / Business Intelligence Consultant
twitter @ToledoSQL
April 29, 2011 at 5:51 am
Thanks for the clarifications everyone. Good conversation.
Chris Umbaugh
Data Warehouse / Business Intelligence Consultant
twitter @ToledoSQL
April 29, 2011 at 6:34 am
A good question and a good explanation.
But why two points, when the numbers were chosen in such a way as to make it almost unimaginable that anything other than the correct answer was correct? If the other minimum had been 30 instead of whatever it was, and the options offered had included 75 and 90 and well as 70, then it might have been worth two points to get it right. I didn't know the algorithm by which the effective maximum is calculated (now I do, so for me it was a very good question and explanation) but it would require a really disfunctional algorithm to produce a wrong answer from that set of numbers. So with the numbers chosen I reckon it would take serious perversity (or a belief in the serious perversity of Microsoft) to choose the wrong option because, after all, using a seriously disfunctional algorithm does indicate perversity.
Arrgh!! I see 36% got it wrong. I hope no-one feels too insulted by the above. But can it be true: 36% of SQLServerCentral subscribers are either seriously perverse or think MS is (or both)? I guess I must be missing something somewhere here.
Tom
Viewing 15 posts - 1 through 15 (of 18 total)
You must be logged in to reply to this topic. Login to reply