Poll: Best book on Oracle Performance/Design/Architecture/Planning

Quote:

---

Originally posted by mikerault
Let's see, I hang around with CEO's, CIO's senior DBA staff, company owners. You all seem to hand around ... here.

Don has his faults, he also provides jobs for a dozen or so consultants, and has helped companies worlwide save money using Oracle. He also devotes many hours to helping the blind with his guidehorse programs (at no profit.) Given the choice between him and what I can see of the stirling personalities here, yes, I would hang with Don.

Each technology needs its, for lack of a better word, cheerleaders, as well as its gadflys. Unfortunately the Oracle world seems to have too few of the former and too many of the latter.

It is always easier to criticize others than to create something yourself, you have no risks, you've put nothing on the line by criticism. Come back when you've written one book on Oracle or given half as many presentations or answered half as many technical questions as either Don or I or even Rich. Only then will your voice mean something. Right now it is just a sounding gong full of fury but no content.

Mike

---

And to think I was complaining that Obfuscation Unlimited was getting boring.

Welcome Mike, and good job. You're going to fit in very well here.

Quote:

---

Originally posted by mikerault
Let's see, I hang around with CEO's, CIO's senior DBA staff, company owners. You all seem to hand around ... here.


Mike

---

WOW, yer neet, can I touch you?

Quote:

---

Originally posted by mikerault
Given the choice between him and what I can see of the stirling personalities here, yes, I would hang with Don.

Mike

---

Uhm, you do realize that this is a contradiction, cause you are chosing to hang with us by posting here.
I can't figure on why though, seems like a lose lose situation to me.
One thing for sure, if you do not have a sense of humor and thick skin, you should definitly NOT post in Obfuscation Unlimited!!

This is a place WE created to blow off steam, bust balls, gripe and have a good time. In here YOU are the new guy, regardless of who you rub elbows with, I would think that would be refreshing.

Wow! Couldn't belive my eyes (as they were still a bit wet from loughing after reading Hanky's welcome to Mike Ault in Obfuscation Unlimited), but yes, it is true: on the top of the forums page it says:

Welcome to our newest member, TomKyte

So, will the battle between the scientists and Burleson Consulting finaly take a more serious note here, in Obfuscation Unlimited? :D

P.S.
Sorry, Tom, to mention your name in a post like this. I just couldn't resist....

Quote:

---

Originally posted by mikerault
Let's see, I hang around with CEO's, CIO's senior DBA staff, company owners. You all seem to hand around ... here.

Don has his faults, he also provides jobs for a dozen or so consultants, and has helped companies worlwide save money using Oracle. He also devotes many hours to helping the blind with his guidehorse programs (at no profit.) Given the choice between him and what I can see of the stirling personalities here, yes, I would hang with Don.

Each technology needs its, for lack of a better word, cheerleaders, as well as its gadflys. Unfortunately the Oracle world seems to have too few of the former and too many of the latter.

It is always easier to criticize others than to create something yourself, you have no risks, you've put nothing on the line by criticism. Come back when you've written one book on Oracle or given half as many presentations or answered half as many technical questions as either Don or I or even Rich. Only then will your voice mean something. Right now it is just a sounding gong full of fury but no content.

Mike

---

Hi Mike,

Do I qualify for that last paragraph? I have answered some questions. I have written some books. I have even seen a database or two (it is strange that Don says he cannot find my "credentials" on the internet? He has my book (at least one -- you can tell him the credentials he was looking for are on the inside jacket flap by the way), I know he knows of my website. I've had dinner with the guy. He tried to get me to write for rampant. Hmm, one wonders why then when interviewing that DBA he didn't remember that?)

Now, I know Don and maybe you find it outrageous that I goto the dictionary to look up terms, but I had to for gadflys (I'm just a Pitt grad with a BA in math after all)

1. A persistent irritating critic; a nuisance.
2. One that acts as a provocative stimulus; a goad.
3. Any of various flies, especially of the family Tabanidae, that bite or annoy livestock and other animals.

I doubt it was #3 you meant?

Well, #1 is a matter of perspective. When you are the one being irritated, sure, seems gladfly like. #2 sounds positive actually. A provocative stimulus -- I like that.

I agree, too many gadflys writing things like your nice ask_tom.htm paper, absolutely agree. (it was nice of don to forward you his emails with all of the talking points -- had a sense of deja vu reading them -- oh yeah, he had already sent them to us...).


If by gadfly you meant "hey, don't criticise our work, we are good people, really" -- sorry, but when someone asks a question, I don't care who's material they pointed to, I try to give an accurate answer to it. If I believe the information to be out of context or character, I ask them to cite a reference and after reading it -- If I find something I believe wrong, well -- I sort of say so. I do it in a fashion such that anyone can then comment on it. If you want to give advice, you better back it up. You can do whatever you want at your customer site, but if you want to print it, back it up.

Take your "indexes love big blocks" paper. http://www.dba-oracle.com/oracle_tip...rge_blocks.htm No caveats, nothing (in an article titled "RULES for setting large blocksizes"). The evidence you point to? a count(*) from a FAST FULL SCAN and you say "that shows indexes like it, for RANGE SCANS", huh. that is trivial to prove "wrong", "not accurate", "not true". (you can try it yourself Mike, in sqlplus, you can prove to yourself that for an index range scan, a block size of INFINITY won't reduce the number of table access by index rowid's, and when range scanning, what contributes the preponderance of the LIO's??)

In a singler user environment.

Conclusively.

Is that a performance proof? Nope, that is just showing how this thing called Oracle actually works

That article boils down to

o indexes love them
o temp digs it
o tpc-c does it, so it must be good

(in fact, that is all that article says). That is the kind of 1/2 truth, misinformation that I'll aggresively respond to with "careful there, be very very careful. sounds all good, but careful with that axe eugene"

Never once have I seen a caveat on index rebuilds -- you keep saying "I say only to do it when it will be beneficial". Well, when is it beneficial and where on dba-oracle.com can I read about when I shouldn't do it?

Making stuff up, just making up numbers, and saying in the article:

Quote:

---

The following stories are true, verifiable accounts (I have witnesses!)

---

Huh, we were supposed to know that the stories are true but the numbers had to be made up. (this goes to the company you keep, you see we don't meet and greet with Don, his private life is just that, his private life. I have a private life too, do good things in my community and places. No ones business but mine and not relevant in a technical discussion forum).

Looking at that paper, I see so many false conclusions from the numbers and facts presented. They don't add up (not just the numbers but the causes and effects). Have you asked yourself how could cursor_sharing=force fix an enqueue wait problem? Or how cursor sharing = force that could maybe help the latch free could have improved performance by 70% when it was 5% of the waits (and the vast -- by far VAST preponderance of waits were enqueue, or something really -- because everything was made up).

So, if by gadfly you mean "these new kids on the block that are pointing out things that are wrong", yes we are here and here to stay. (I hardly consider myself new however, you are not that much older than me, but you refer to us as such, your "careful what you prove" thing:

Quote:

---

So now we have a new crop of experts providing proofs (sometimes limited to a single-user, small database on a laptop)

---

(oh yeah, I've spent my fair share on hardware. I had to insulate my "computer room" downstairs, the fan noise was keeping my mother in law awake from the servers. I use a laptop on the road -- but you know what, the same numbers pop out on the poweredge in the basement....)

sign me,

sincerely
a. gadfly

Hmmm, I hadn't seen Mike's article referring to Don's article referring to Robin Schumacher's book (which I'll not be splashing out my cash on, btw), but it seems to me that the phrase "Note the test that proves it" contradicts everything that Mike (and Don) has said about scripts not proving anything.

I'm tired of making the point that as far as I can see, the much-lauded TPC-C tests that I've seen referenced use large block sizes only for clusters, which makes some sense to me. So I agree with Mike: "These vendors spent hundreds of thousands of dollars to get the fastest performance and their choice of multiple blocksizes speaks for itself." ... and they didn't use them for TEMP space, not for INDEXES. Unless I've misread the gloriously obfuscated scripts.

This must be about the fourth time I've pointed this out to the ranks of Burleson Consulting, and in the abscence of an answer I'll just conclude that they're too embarrased to acknowledge that I've found a fatal contradiction in their arguments. Otherwise they might like to forward Robin's script to HP, UNISYS and Oracle to let them know what an opportunity they're missing.

brilliany, love it :)

you can do as much kicking screaming as you want Mike - but the more you do the small bit of credability you have goes away.

Quote:

---

Originally posted by TomKyte
1. A persistent irritating critic; a nuisance.
2. One that acts as a provocative stimulus; a goad.

---

Welcome to DBA Support.com. I would consider myself one and two but not three.
If you stay around think definitely won't get boring. :)

Quote:

---

Originally posted by mikerault
Let's see, I hang around with CEO's, CIO's senior DBA staff, company owners. You all seem to hand around ... here.

---

Mike, it is true that I don't hang around CEO's and CIO's, but I was confused on one issue concerning your TPC Benchmark Example Look at page 379 for reference.

I found this link from the article where you talk about using large block sizes, yet in one of your benchmarking examples almost all of the block sizes along with the default block size are 2k and in the other the default block size is 4k. I thought 2k block sizes went out with 7.1. 4k isn't very large either.

It seems that from a management perspective it would be easier to use 8k for OLTP systems and 16k or 32k for datawarehouse systems. I realize that this approach is somewhat simplistic and certainly not all inclusive. But I'm not sure how using large block sizes in OLTP would help since most transactions would probably be small, most of the current data would be cached anyway, and if it weren't the OS is likely to grab multiple blocks on each read potentially causing Oracle to read more than it needs. It seems to me that the basic theory of your article is flawed, and would be seen as such to most DBA's.

Can you explain why an article talking about using large block sizes is referencing a benchmark test where they are using small block sizes???

Quote:

---

Originally posted by TomKyte
Making stuff up, just making up numbers, and saying in the article:

"The following stories are true, verifiable accounts (I have witnesses!)"

Huh, we were supposed to know that the stories are true but the numbers had to be made up.

---

Regretably we learn that the stories are made up too:

Post by DaPi:
http://dba.ipbhost.com/index.php?sho...indpost&p=5347

Quote:

---

Since I spend my week "sitting and thinking" I've only just had time to look at all of JL's article: http://www.jlcomp.demon.co.uk/review_01.html

At the end of the article is DB's e-mail reply to JL's comments on the fabricated statspack reports:

Quote:

---

- This obvious fabrication makes the author look ignorant and incompetent;


Actually, yours is the first complain I've received. As you noted, even a beginner can see that the examples are "obvious fabrications", designed solely to illustrate the points.


This is one of the most frequently-read DBAZine articles they ever published, and the huge amount of positive feedback prompted me to create the book "Oracle Silver Bullets".


http://www.rampant-books.com/book_20...ver_bullet.htm


Since, you acknowledge that they are obviously nothing more than illustrative examples, you have a no basis for labeling them "ignorant and incompetent". This remark is incorrect, rude and unprofessional.

---

In the "Oracle 911" article, http://www.dbazine.com/burleson20.shtml DB presents two of these statspack reports as part of "war stories", in Canada and Germany, about DB's successes in using his silver bullets. Are we to conclude that these stories are also "obvious fabrications, designed solely to illustrate the points"?

---

Post by burleson:
http://dba.ipbhost.com/index.php?sho...indpost&p=5354

Quote:

---

Quote:

---

In the "Oracle 911" article, http://www.dbazine.com/burleson20.shtml DB presents two of these statspack reports as part of "war stories", in Canada and Germany, about DB's successes in using his silver bullets.

Are we to conclude that these stories are also "obvious fabrications, designed solely to illustrate the points"?

---

Er, yes, everyone seems to understand this:


I made it very, very clear in the article:

Quote:

---

To those scientist DBAs who demand “proofs” that these methods worked, you are not going to find it in this article.

---

---