November 12, 2010 at 8:22 am
GilaMonster (11/12/2010)
Stefan Krzywicki (11/12/2010)
Once everone who went to PASS is back (I think it is over now, yes?)No, today's the postcon day (should be starting in about an hour).
Got it, thanks! I shall hold my horses for at least one more day. : -)
--------------------------------------
When you encounter a problem, if the solution isn't readily evident go back to the start and check your assumptions.
--------------------------------------
Itβs unpleasantly like being drunk.
Whatβs so unpleasant about being drunk?
You ask a glass of water. -- Douglas Adams
November 12, 2010 at 8:23 am
GilaMonster (11/12/2010)
I swear this project is going to drive me to drink.Email from BA/architect:
Urgent! Business confirms that it is possible to have more than one X per Y. Out design must cater for this.
My reply:
During my initial analysis I identified that this was the case based on a discussion with <someone>. The current system design allows for any number of X per Y and has done so since the beginning.
Should have replied back with "You sound like a bunch of panicking students who don't know how to do their assignment and subsequently post it in the forums for somebody else to do."
Would that have gotten the BA's goat?
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
November 12, 2010 at 8:23 am
Jeff Moden (11/12/2010)
Paul White NZ (11/11/2010)
Did you also bust the myth that table variables are always estimated as containing one row?Statement level recompile?
Yup.
There's still no statistics, but the storage engine knows the total rows in the table. The optimiser can generate a non-1 estimate from that. It may still be totally wrong, because there's no density information, no histogram, so no way of telling what the data actually looks like. The estimate appears to be a 'magic number' formula based just on the total row count in the table.
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
November 12, 2010 at 8:26 am
CirquedeSQLeil (11/12/2010)
GilaMonster (11/12/2010)
CirquedeSQLeil (11/11/2010)
I have a few articles that I am finishing up currently.Topics?
Physical Join Operators
Toss it over. If I need a break from genetic algorithms this weekend I'll take a look. Will just be brief comments.
gail (at) <blog domain>
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
November 12, 2010 at 8:27 am
The Dixie Flatline (11/11/2010)
Congratulations, Wayne. Applause! Thunderous applauseLet me ask what logging takes place with table variables, since they are not affected by rollbacks.
They are not affected by explicit rollbacks (begin tran / rollback tran). They are affected by implicit transactions (update/insert/delete statement failing, causing a rollback of all previously changed rows by that statement).
Please refer to Gail's blog http://sqlinthewild.co.za/index.php/2010/10/12/a-trio-of-table-variables/ for everything you need to know about that - it explains it very thoroughly.
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
November 12, 2010 at 8:28 am
Paul White NZ (11/11/2010)
WayneS (11/11/2010)
I started the presentation asking questions. Everyone thought that table variables were 1. memory-only, 2. did no logging, and 3. had no indexes. Now, about 20 more people know better.Well done. It's amazing how few database people know the basics of table variables.
Did you also bust the myth that table variables are always estimated as containing one row?
I assume you mean with the OPTION (RECOMPILE) hint? Sure did.
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
November 12, 2010 at 8:29 am
GilaMonster (11/12/2010)
I swear this project is going to drive me to drink.Email from BA/architect:
Urgent! Business confirms that it is possible to have more than one X per Y. Out design must cater for this.
My reply:
During my initial analysis I identified that this was the case based on a discussion with <someone>. The current system design allows for any number of X per Y and has done so since the beginning.
I think that particular project might drive me to felonies...
---------------------------------------------------------
How best to post your question[/url]
How to post performance problems[/url]
Tally Table:What it is and how it replaces a loop[/url]
"stewsterl 80804 (10/16/2009)I guess when you stop and try to understand the solution provided you not only learn, but save yourself some headaches when you need to make any slight changes."
November 12, 2010 at 8:31 am
CirquedeSQLeil (11/11/2010)
WayneS (11/11/2010)
WayneS (11/11/2010)
Well, I'm off to give my first presentation of "Comparing Tables Variables and Temporary Tables" to my UG tonight. Wish me good tidings, smooth talk, and all the other stuff you need for public speaking.Well, I think it went pretty well. In a similar vein to Gail's "statistics" 5 times in a minute, I did get a little bit tongue-tied saying "tempdb" and "temporary tables" in the same sentence. I ran the PowerPoint presentation in presenter mode, but then ran into some minor issues flipping back and forth with SSMS and my virtual machine back to the slide. I might try setting up the monitor in slave mode the next time.
I started the presentation asking questions. Everyone thought that table variables were 1. memory-only, 2. did no logging, and 3. had no indexes. Now, about 20 more people know better.
Congrats and good job.
When can I book you to speak for my User Group - same topic?
Almost anytime - though we need to ensure a better internet connection than what I ran into trying to view your last meeting.
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
November 12, 2010 at 8:34 am
Henrico Bekker (11/12/2010)
another one....http://www.sqlservercentral.com/Forums/Topic1019796-1550-1.aspx?Update=1
oh and I reached a 1000 point...small milestone π
Congrats on the millennial mark!
- 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
November 12, 2010 at 8:36 am
GilaMonster (11/12/2010)
WayneS (11/11/2010)
Everyone thought that table variables were 1. memory-only, 2. did no logging, and 3. had no indexes. Now, about 20 more people know better.And I thought the 'no logging' myth was a rare one. Glad I did spend some (lots of) time debunking that one.
Not only am I glad that you did so, but also that you let me use your work in my presentation. THANK YOU!
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
November 12, 2010 at 8:36 am
Because you're bored, it's Friday, and I'm stuck :-D:-P
http://www.sqlservercentral.com/Forums/Topic1019985-1633-1.aspx
---------------------------------------------------------
How best to post your question[/url]
How to post performance problems[/url]
Tally Table:What it is and how it replaces a loop[/url]
"stewsterl 80804 (10/16/2009)I guess when you stop and try to understand the solution provided you not only learn, but save yourself some headaches when you need to make any slight changes."
November 12, 2010 at 8:37 am
The Dixie Flatline (11/12/2010)
Paul, thanks for the reply. I understand now.
Okay, that's 21 now! :w00t:
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
November 12, 2010 at 9:02 am
GilaMonster (11/12/2010)
I swear this project is going to drive me to drink.Email from BA/architect:
Urgent! Business confirms that it is possible to have more than one X per Y. Out design must cater for this.
My reply:
During my initial analysis I identified that this was the case based on a discussion with <someone>. The current system design allows for any number of X per Y and has done so since the beginning.
How about - can you test and verify I accounted for this correctly?
I thought this was covered very early on.
Must be the Urgent! that rattled you.
Greg E
November 12, 2010 at 9:02 am
Henrico Bekker (11/12/2010)
another one....http://www.sqlservercentral.com/Forums/Topic1019796-1550-1.aspx?Update=1
oh and I reached a 1000 point...small milestone π
congrats!
November 12, 2010 at 9:06 am
GilaMonster (11/12/2010)
CirquedeSQLeil (11/12/2010)
GilaMonster (11/12/2010)
CirquedeSQLeil (11/11/2010)
I have a few articles that I am finishing up currently.Topics?
Physical Join Operators
Toss it over. If I need a break from genetic algorithms this weekend I'll take a look. Will just be brief comments.
gail (at) <blog domain>
Sent - thanks for the help. Brief comments are good. π
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
Viewing 15 posts - 21,421 through 21,435 (of 66,738 total)
You must be logged in to reply to this topic. Login to reply