« first day (1303 days earlier)      last day (3551 days later) » 
01:00 - 20:0020:00 - 23:00

1:07 AM
SOFA KING SICK OF ALL OF THIS CRIPPLEWARE ON MY LAPTOP!!!!!!!!!!!!!!
 
 
2 hours later…
3:32 AM
@swasheck so delete it all?
 
@PaulWhite I've modified my answer, and hopefully, incorporated all the salient points. Thanks again for the discussion - I know your time is very valuable.
2
A: SQL Server: Create New GUID Values for Existing Index

Max VernonYou may be able to reduce index rebuild/reorg times by moving to NEWSEQUENTIALID(). You are currently inserting around 4,500 rows into the table per day. The table has around 6.3 million rows at present. That's around 0.7% of the table per day. Assuming every insert on the table results in an...

 
 
3 hours later…
6:24 AM
How about moving this one over here?
0
Q: How to start an sp_procoption Stored Procedure without restarting SQL Server

Critical ErrorIs there a way to start a stored procedure that has been configured for automatic startup via sp_procoption without having to restart SQL Server? Occasionally I'll need to make an adjustment and have to KILL the running session to do so, but I cannot seem to locate a way to turn it back on agai...

To me it seems in the grey area but I wonder if our dear experts have an alternative to what Damien suspects is the only way and if they'd be more persuaded to post an answer if the question was on DBA.
 
I think so, voted
Happy Birthday @jcolebrand :)
5
 
 
5 hours later…
11:17 AM
Kind of interesting question about two SQL-Server queries:
0
Q: Why is the clustered index seek much slower on SELECT with OR condition compared with single separate SELECTs

VDohnalBased on theese questions and the answers given: SQL 2008 Server - performance loss possibly connected with a very large table Large table with historical data allocates too much of SQL Server 2008 Std. memory - performance loss for other databases I have a table in a database SupervisionP def...

Is it a statistics issue or the optimizer can't find the best plan for SELECT min(cas) from PenData p WHERE IDUkazatel=24 OR IDUkazatel=25 ; ?
Would rewriting as SELECT min(cas) from PenData p WHERE IDUkazatel IN (24,25); make any difference? (I guess not)
Adding GROUP BY and TOP 1?: SELECT TOP (1) min(cas) AS min_cas FROM PenData p WHERE IDUkazatel IN (24,25) GROUP BY IDUkazatel ORDER BY min_cas;
 
 
1 hour later…
12:34 PM
@ypercube It's a rather sad limitation. The optimizer doesn't generate an index union strategy for disjunctions by default. It can do this if FORCESEEK is used, but not if the index forced is the clustered index.
Actually it might require two different indexes, I forget the details offhand.
 
@PaulWhite That user has irritated me. Check their comments and their self-answer (after they accepted David Spillet's answer!)
 
@ypercube I can only guess he has many items in the IN list.
 
@PaulWhite No problem, thnx. I was just curious anyway, I rarely use SQL-Server nowadays. I cannot even test at work, except with SQL-Fiddle
 
@ypercube What's your engine of choice these days?
 
We use Postgres in the work I am now.
 
12:53 PM
Thanks. I added a minimal answer, but I may regret it.
Thanks, this is what I was looking for... — VDohnal 19 secs ago
^ Saints be praised.
 
@PaulWhite :) It's amazing how many tricks/uses there are for CROSS/OUTER APPLY
 
Does PG call it a lateral join?
 
Yeah but I find the syntax weird.
It was only added in the latest version.
 
I think, with hindsight, I would have preferred CORRELATED JOIN...USING or something like that. I prefer apply to lateral join though.
 
1:16 PM
Not sure how to write a clear question but I really struggle with bulk importing a InnoDB Table with ~10 million rows (~ 7GB). I did some research about various options inno_buffer_* but then again testing is very time intense and I feel I'm missing some clear direction to move ;C
In the moment I use different dumps with n million rows and stop time importing but also dropping the database takes > 1hour so yes - testing process becomes very slow
 
@yoshi Do you have the data in csv files?
You could use LOAD DATA INFILE
 
@ypercube: Yes I use CSV with head row (column names) - and LOAD DATA LOCAL INFILE
 
I think LOAD DATA INFILE is better for MyISAM tables.
Perhaps you can load into a myisam table and then copy the data to the InnoDB one.
 
fyi this is the basic setup I use pastebin.com/tQdeUP5N
Hmm... afaik InnoDB has better support for FK keys (or generally has supoort as opposed to MyISAM) - have you ever done something like this? I wonder if I would lose information then
 
All that detail seems like a good basis for a question to me. Just keep it focussed.
 
1:31 PM
@PaulWhite You think so? It feels like a "here's my config tell me what to do question" … though that's exactly my problem :D
 
@yoshi Sounds like a "how to I optimize bulk inserts to an InnoDB table" question to me :)
But then I know nothing of MySQL, so I might be wrong.
 
@yoshi I mean you use the MyISAM as temp only
 
@ypercube The CHANGE ENGINE -command (have to google it) just wondering how well conversion works. Anyway I will surely try it out after my DROP TABLE statement has finished (1h and counting ^^)
@PaulWhite Perspective :) But yes why not - I'm phrasing a question atm
 
2:07 PM
dba.stackexchange.com/q/72750/42199 critic comments? any suggestion to improve question quality? :)
 
@yoshi Add the RAM of the machine.
 
helo
just wanna ask something. anyone up here?
 
JNK
yes
 
how do u think about Big Data?
i just dont know why BigData related of mapreduce
and why that noSQL related with BigData
 
JNK
BigData is a buzz word
6
 
2:15 PM
buzz?
 
JNK
unless you are google or facebook or twitter or the NSA you don't have big data
You need to be more specific if you have a question :)
 
oh ,
 
JNK
Yes Big Data is a term that doesn't really mean anything
people use it a lot in articles and whatnot
But your 2 TB database is not "big"
 
wow 2TB is not big?
 
JNK
not really
It's big to people who misuse the term "Big Data"
 
2:17 PM
yeah many people like me mislead about that
 
JNK
it's not your fault it's something people talk about a lot
but most serious data professionals don't
 
yeah
thanks for make me understand
 
JNK
so anyway, NoSQL is good if you don't care about consistency of your data
for social networks and whatnot it's great
 
i just dont familiar with NoSQL
 
JNK
in those scenarios it's not mission critical that you see Aunt Alice's casserole recipe right away
for business use relational databases are usually a better fit because it DOES matter if you are looking at the most recent version of the accounts table
There's posts about it on several SE sites
You're basically trading consistency for sharding
 
2:19 PM
where?
 
JNK
stack overflow, DBA.SE (which is the site this room is for)
 
yeah, in collage i learn bout SQL
and when i see NoSQL i just a little bit confused
why is that can be modeled? btw?
 
JNK
NoSQL uses something called eventual consistency in most implementations
 
what kind?
 
JNK
so in a SQL DB nothing can read data until it's written
in NoSQL you will get the updated data at some point but not necessarily right away
 
2:21 PM
u mean its all about clustering?
 
@user44517 If you can walk into a store and buy a computer with a hard drive that can hold all your data, it's not big
 
& yeah JNK explained that
 
But I like to talk. It makes me feel important
 
haha yeah
for SQL itself any method like DBMS Design are exists,
seems like ORM, that provide abstraction in DB to Object and Object to Real Life
but what about NoSQL itself? is it can be modeled with ORM too?
i just wondering for mapred itself, caching technique are the tecnique for make it faster
so as far as i can understand, NoSQL leave all of the Searching Algorithm , Indexing and formating data to developer , is that true?
 
2:38 PM
@yoshi do check this answer of Rolando's
might be useful
 
@ypercube thanks! I've updated the question (8GB ram over here)
@Marian thanks: I'll have a look!
 
where do u get the latest information about database technology ?
@yoshi wow server with 64GB?
what kind?
 
But it executed in SSMS.. — Niks 4 mins ago
Someone spare a fist of death? Mine's in the shop
 
2:55 PM
Respected cube?
 
yay. I was selected for SQL Sat Denver, but we still have to iron out which presentation.
6
@jcolebrand can't. it's corporate crippleware
 
@swasheck Not Tom?
Oh I see. You were selected, just not that session.
 
@swasheck Congrats!
 
@user44517 64GB? where? must be a typo o.O
 
@yoshi dba.stackexchange.com/questions/20372/… the answer he gave to u
after i google it, there are some server with 64GB, :\ well he share some knowledge
 
3:03 PM
@PaulWhite so i replied with "yes. i'd much rather to the stats management preso"
 
@billinkc I didn't even think that SSIS had a query builder.....
 
ehrmagerd!! today is @jcolebrand 's boithday
 
happy birthday @jcolebrand!!!
5
 
@user44517 Oh I see :) Well just after "I've a 10 cpu box" I was like "grml rich kid" - nope I do not have 64GB … actually why?! :)
 
@JNK big data is more than a buzzword, it's a highly subjective concept that is loosely translated, "i'm pushing the limits of my experience/competence in tuning this thing."
 
JNK
3:06 PM
@swasheck yeah I guess it's sort of like "Fast car"
a fast car to me would not be a fast car to Mario Andretti
 
mmhmm.
 
@yoshi your edit regarding the storage gets me thinking that's your disk it's the problem; are you reading and writing from/to the same disk?
 
@yoshi that must be awesome server 64GB , i want one after i check theprice its a 1659 $
 
@yoshi my kids love your island
 
3:10 PM
1649$ is my 4 month salary :\
 
@user44517 that's terribad
 
@JNK did you not like my fast car reference?
 
well well well :p
 
@Marian: Yes I actually do. As a minimal precaution I try to avoid excessive i/O operation (where possible…) not sure how it would be on our dev sever but for my working notebook ... you think external hard disk could speed up.
 
@Kermit that's not nice
@user44517 where do you live/work?
 
3:12 PM
Oh also the hard disk is encrypted (this default thingy in ubuntu 14.04 => full disk encryption) this could be problematic?
 
@swasheck that was my next question
 
indonesia,
 
oh
that's not terribad than
that's averagebad
 
haha yeah
its all about living cost
 
my colleague's wife is from indonesia
 
3:12 PM
@swasheck Of course they do 8) Actualyl we try to get officially recognised by UN so we could live from tourism and taxes instead of writing code (these human keyboards are hard to use with dinosaur pawns :3 )
 
he said he could have 4 servants and a lavish home for $500/month
 
haha yeah , so cheap in here
 
> Just the other day I heard a soldier falling off some Indonesian junk that's going round.
 
what kind?
 
song lyrics ... Surrender by Cheap Trick. old old song
 
3:14 PM
wow
i;ve never heaard that :p
 
> Mommy's alright, Daddy's alright, they just seem a little weird.
@user44517 dont know if serious or not
 
idk , whats meaning of that lyrics
 
i dont know either
 
whatever
 
seriously. no clue
people have thoughts
 
3:18 PM
if that means indonesian junk , i must admit yes
people here in passes collage so easily
umm silence
 
@swasheck srsly guyz, srsly
 
@yoshi single encrypted notebook disk? I'd reckon that's your bottleneck. That hybrid (ssd) part is not really useful for bulk writes.. so you're actually reading/writing on a slow disk.
 
> You guys, you guys! Oh my god, seriously, you guys!
@yoshi hard drive encryption is the bane of my working existence. it takes a really nice laptop and turns it into a piece of junk
 
@jcolebrand doesn't seem like a great question... period
 
belongs on devnul.se
 
3:26 PM
BOOM!
 
Yes, not a big fan of encryption either but my boss has made it mandatory.
The very specific table in question contains some not-exactly-uninteresting data about various people and their political views so I can understand they dislike the idea of having some dump getting lost in the wilderness ~ which I must confess is quite likely in my life :D
 
@yoshi in which industry do you work?
 
is NGO an acceptable answer?
 
sure
@PaulWhite ... OPTION (QUERYTRACEON 2312); causes SQL Sentry Plan Explorer to timeout before collecting a plan. THOSE are the kinds of queries with which I'm working
 
So I guess that's also why our server are not best thing money can buy. I was thinking even though I might optimise import speed and can archive some data table will still grow and the problem with a backup server remains (to get it ready in short time).
So i wonder if there's something like incremental backups … ? more like an idea but if someone could suggest in which direction to dig
 
too easy, i see :d
 
@ypercube :*
 
is the ansi name windowing functions?
0
Q: Formal name for functions that use the OVER clause

Joel CoehoornIs there a formal/academic name for the family of functions in Sql Server like ROW_NUMBER(), LAG, LEAD, RANK etc, that make use of the OVER(ORDER BY <PARTITION BY> ) syntax? How much of this is covered by the ansi standard? I can see on MSDN that some of those are listed as analytic functions, b...

 
@bluefeet I think yes.
> Window functions
A window function is a function whose result for a given row is derived from the window frame of that row as defined by a window structure descriptor of a windowed table. Window functions may only appear in the <select list> of a <query specification> or <select statement: single row>, or the <order by clause> simply contained in a <query expression> that is a simple table query.
> A window function is one of:
— A rank function.
— A distribution function.
— The row number function.
— A window aggregate function.
— The ntile function.
— The lead function.
— The lag function.
— The first-value function.
— The last-value function.
— The nth-value function.
 
3:44 PM
-1
Q: How to display money in 12 Lc or 2 Cr

Surendrahi i want display money as 12Lc or 2Cr if money =1200000 as 12 Lc when 12000000 as 1.2 Cr how can i do this help me guys

 
@bluefeet know of a deer valley rd?
 
@bluefeet Feel free to add the list (or the whole text) in your answer.
it's from the SQL docs (draft)
 
Yes that would be a good edit
 
I'm @paulwhite and I approved this edit
 
:( I didn't mean it to sound that way
 
3:57 PM
@swasheck why yes I do
@ypercube I was trying to find some reference material the answer
 
well i like this chat btw, many professional here
 
@ypercube I added that list along with a link to site with more details -- simple-talk.com/sql/t-sql-programming/window-functions-in-sql
 
@bluefeet I have in this answer, a link to wiscorp.com page, where they have several drafts of the SQL standard.: dba.stackexchange.com/questions/28698/…
 
@bluefeet a recruiter just emailed me about a client who's looking for a dba
for a client on that road
@PaulWhite it was a play on the political ads here in the US
 
Oh! Ok then :)
 
4:04 PM
@PaulWhite this awful query (plan)? ... still compiling the estimate after 28 minutes
 
@swasheck Excessive compilation time is a 2014 CE bug I've seen before.
 
@swasheck fun, that is pretty far north from where I live
 
I forget the exact plan shape that triggers it, but it might have been lots of hash joins.
Not fixed yet AFAIK
 
that's good to know. it's only with the new CE, btw
 
Yes, it's specific to cardinality estimation in the new CE. It gets a bit recursive.
 
4:08 PM
well this query is truly the stuff of nightmares. seems Cognos has quite the capacity to "invent a better idiot"
 
what is CE?
 
Cardinality estimation/estimator
A component of the SQL Server query optimizer
 
oh is that somekinda project manager task ?
oh i see
 
"how many rows do i expect this (and each) phase of the query to return"
 
ok go on
 
4:10 PM
@user44517 i've tapped my knowledge. @PaulWhite is the guy to whom you should listen :)
 
yea. thanks :)
 
nice
-2
Q: Best website for Microsoft Certification Dumps

user3789498Can someone please assist me with a website to download microsoft certification dumps. I was checking the website www.braindumps.com but they do not seem to offer the actuall exam questions Thanks

> And how will they find out if i cheat
 
but my eyes cant hold it, and also there are the beattles tribute concert are shown on TV now. so i have to sleep , its 12 midnight in indonesia
 
quality user right there
 
so i am gonna sleep
thanks for chat i will chat again next time
 
4:15 PM
hello hello, i am searching for one of my question in stackoverflow. Question is related to sql information_schema about retrieving table, column and other details. I had answers posted. This was sometime 3-4 months ago. I couldn't find my own posting. Whether my question got deleted or how should i able to retrieve my question ?
 
@bluefeet can you not delete that q.. i wanna submit a dba reaction
 
@Kermit I'm not
@goofyui You should be able to see recent deleted questions from your profile
 
Thank you folks. I am searching on it
 
@goofyui You can sort by date (click on "newest")
 
4:27 PM
Thanks @JackDouglas @bluefeet @swasheck (I think I got the transcript right on those wishes)
 
Is there anyway for me to sort by keyword such as Information_Schema
 
When you visit your profile, the search box is prepopulated with your user id. Keep that there and then add in your search terms
 
If keyword sort is missing, That would be hardest part. I may have to repost again the same question. This something Stackoverflow has to fix it
 
@goofyui there's no result...
you can custom search by using user:620720 information_schema
 
@goofyui you can use stackoverflow.com/…
to search through your questions
 
4:30 PM
There is something called Advanced Search : stackoverflow.com/help/searching
 
What permission is needed to run dbcc freeproccache?\
 
@ypercube Import is no dobut faster with MyISAM problem comes when changing engine back to InnoDB.
However answers have pointed out the keys might be a problem.
Thanks again for you help!
 
@Zane the expressed written consent of major league baseball
 
@yoshi Yeah, I saw. You have like 13 indexes on the table.
 
4:44 PM
yes... well... sounds like much, not? ^^;
Sorry for confusion - I'm quite new to this "big data" stuff :> (read: I haven't designed the database nor do I [after 4 weeks] know why one would do that + the sheer amount is kinda blasting me)
 
@swasheck I got the email that says I'm speaking in Denver. Will you be after-partying or lame as usual?
 
@mmarie somewhere in between
 
cool
 
yeah
 
Now I have to start building my half day session for SQL Sat Denver.
 
4:55 PM
@PaulWhite Thanks!
 
No, I just downvoted. Started to comment and walked away
 
@billinkc :D
 
And the answer sounds like DV material. I need a VM to test the assertion that the 32 bit version isn't installed by default
 
@MikeFal what're you doing?
 
5:04 PM
@swasheck Introduction to Powershell for the SQL DBA. You'd like it, more of that no-holds barred scripting.
 
cool
 
SCRIPT ALL THE THINGS
 
I have to give a presentation tomorrow, any bets on how many times I'll get sick before it?
 
7
 
Why would you get sick?
 
5:06 PM
@MikeFal nerves
 
Psh
 
@MikeFal not done a technical one before
 
Nothing to be nervous about. People there WANT to hear what you have to say.
 
@bluefeet just get up there and do it. i go numb and incoherent
 
Own that shite.
 
5:07 PM
@MikeFal @bluefeet exactly. it's like @billinkc told me, "nobody's done the research you've done. nobody's done the work you've done. they're there to learn from that."
 
+1
Then next year, you will submit to SQL Saturday Denver and come share with us.
 
/jedi hand wave
 
well if it was a topic I was more comfortable with than I think I'd be ok. It's still not a topic I'm great at so we will see how it goes
 
still parameter sniffing?
 
@swasheck yes
 
5:12 PM
smells like sulfur
 
Smells like poop
 
@swasheck you are taking advice that doesn't involve drinking or wearing a skirt from @billinkc??
or maybe it did involve all of that
 
@Lamak yeah. he was wearing a skirt when he told me that.
 
and you were drinking?
 
But I wasn't trying to get you drunk, yet
 
5:29 PM
 
wtf????
 
@bluefeet what kind of question would that site get?
 
Probably a lot of computer and role playing gamges based ones
2
 
it's been removed
 
JNK
The last slide of your presentation should sum up ALL its key points.Like this one, from our company-wide town hall: http://twitter.com/StackExchange/status/494537484093489153/photo/1
 
5:45 PM
carp question?
0
Q: DB table design

anilWe are building a database for a website which is related to the farmers land certification and storing the related data. Table : farmer_detail(id,name,address...) we store farmers details who register their land. After that there is a form which is as below: Sl_no Details of operation ...

 
6:12 PM
@MaxVernon that's a big fat huh? question
 
@bluefeet yup
I like to think I'm pretty good at understanding non-English speaker's questions, but not this one!
 
6:25 PM
what's a "Hard Page Fault" in Windows? i tried reading about it and i think i'm not quite sure. basically, it's a memory page that Windows doesnt have available so it has to go get more?
 
That's my approximate understanding as well
 
6:54 PM
thanks william
 
Why am I in trouble?
 
are we using formal names now?
 
Do you have a problem with that AzureFeet?
 
glad no one knows my name
 
@bluefeet I know it :)
 
6:54 PM
shhhhhhhh
 
as do i
 
This must be the worst secret known to mankind
 
haha
hush all of you, just hush
 
is it ... uhhhh
Karen?
 
no
 
6:56 PM
and I was about to link to the times it's been mentioned in chat
 
Sharon?
 
@billinkc the docs certainly do say so
 
@swasheck nope
 
Meryn?
 
you lucky folks in the heap are the only ones who know my real name
 
6:57 PM
not even your spouse knows your real name?
 
Wow that presentation was terrible.
 
@swasheck aaaw, you beat me to it
 
honesty and communication are the foundations of a healthy relationship, @bluefeet
@Zane were you a witness or a perpetrator
 
@bluefeet I know it as well. I've invited you into the order of peoples with kick ass names.
@swasheck perpetrator.
 
@swasheck :)
 
7:00 PM
@Marian I tip my hat to you and in 23 hours, I'll reward/punish you with SO rep
 
Demonoid is back online
 
@Lamak sarcasm is the only thing at which i can beat you.
 
It turned into a less of a execution plan presentation and more of a free form SQL server question and answer session.
@Kermit huh.
 
(and stilted english)
@Zane did you just say, "NO FURTHER QUESTIONS!! GFY!!! STAY ON TOPIC!"
 
7:01 PM
@swasheck no I just answered questions about SQL server for about an hour an a half.
@Kermit Blocked.
 
@swasheck hey, you know your ways into being a DBA, you also have experience doing presentations in the SQL Server community
 
it started with execution plans and by the end of it I was making props to explain how pages are pulled.
 
seems more of an attendees knowledge level issue
 
@Lamak no. i'm just not afraid to look stupid by asking stupid questions
 
@billinkc thank you, I do need points for down votes (but for a period I'm safe).
 
7:08 PM
click here ... what do you see?
@Zane what kind of props?
 
@swasheck pages with scribbles representing data.
 
yeah, no up-votes, exactly like Bill :)
 
@billinkc odd. it told me it was waiting for moderation
 
Two questions enter, swasheck's leave
 
7:10 PM
looking at the auction site for a projector there is a vivitek home theater projector in there for 1000 dollars off.
 
GTFO
 
1500 technically.
 
do i hear 2500?
 
Have I already asked you people what online SQL formatter you use?
 
I don't
 
7:12 PM
 
@Kermit that's the one i use
@Marian i've used that one in the past as well
the queries that cognos generates generally overlow the capabilities of these tools
 
7:24 PM
@Marian ty
 
7:41 PM
@swasheck :). Same for EF ugliness.
 
@Marian just for fun i copy-pasted a cognos query that i captured into Word. it was 23 pages.
 
@Kermit no problem. Though if it's inherited code I prefer to format it myself while reading it.. so I can understand it better.
@swasheck wow, my longest generated query from EF was 3-4 pages. You rock.
 
i rock for being able to maintain my sanity?
 
yep
and trying to understand that :)
 
but cognos is the bees knees. i mean, the people who wrote the ORM under it all worked on the optimizer teams for sql server, oracle, and db2
at least that's what the local cognos evangelist tells us
 
7:45 PM
I see.. so did you ask the evangelist to help you tune those queries?
maybe he worked for @PaulWhite ..and taught him some stuff :).
 
yeah. he said it's an ugly query
->rage<-
 
Hooray for autogenerated procs. We have the SelectAll, SelectByPK, SelectPage and SelectPageByCriteria procs
But wait, there's more!
 
@swasheck ugly? 23 pages of SQL is art.
 
We also have the required views for the table which is ... pretty much SELECT * FROM table
But we of course also have SelectAll, SelectByPK, etc for the view as well
 
@billinkc ALTER TABLE [table] ADD [dont_you_date_select_star] as 1/0;
@Marian a "modern art masterpiece"
 
7:51 PM
@swasheck They at least enumerated the columns. We're getting rid of 7 guids in one table but I get to find all the places it's used
 
@swasheck "un chef d'oeuvre"
 
Which is a perfect use case for SSDT, if only the model would validate
 
hahahahah. that sucks
 
01:00 - 20:0020:00 - 23:00

« first day (1303 days earlier)      last day (3551 days later) »