« first day (1525 days earlier)      last day (3642 days later) » 

10:56
be merged with the identical at DBA?: dba.stackexchange.com/questions/94545/…
@bluefeet, I've flagged the SO one.
Or is it a problem that both have an accepted answer?
I just migrated the SO on here, and VtC as a dup. A mod will have to merge it
thnx
11:46
Clickety
JNK
JNK
12:40
morning all
13:26
morning
13:49
@Nico what are you talking about? Can you prove your claim that ISNULL() is faster than COALESCE()? — ypercube 53 secs ago
His profile show he's a SQL-Server guy. I bet he did not notice it's a mysql question.
@ypercube doesn't ISNULL in MySQL returns 1 or 0?
@Lamak No, it returns an error. IFNULL, on the other hand...
@AndriyM ah, I shouldn't talk about anything MySQL related, my ignorance shows
@Lamak Yes, you are right.
Ah, you see, @Lamak, it was me who was ignorant :)
13:56
well, look at that
Thnx Lamak, I commented about that. I rarely use ISNULL and IFNULL and NULLIF, for that exact reason
(that I have to look up what each one does every time)
@ypercube yeah, I just vaguely remembered about ISNULL on MySQL
@ypercube ISNULL() can be faster than COALESCE() in one scenario: mssqltips.com/sqlservertip/2689/…
@AaronBertrand Yeah, I remember you had a blog about it. The comment was on a mysql question ;)
Kill:
-1
Q: How redo is transfered to standby database?

nvsnreddyAs we all know that redo is transported from primary db to standby db .By referring diff docs I came to know that redo data is transfered by i)reading from redo log buffer in one of the modes(max availability/protection/performance) , ii)redo logs are transfered in other mode and iii)archive logs...

Can't find his previously closed question that he posted yesterday
14:07
He self-deleted
14:45
@MikaelEriksson - nice answer dba.stackexchange.com/a/94697/10832 - to bad the OP did not ask his question properly.
15:01
@AndriyM You were right, too ;)
EVERYONE GETS A RIBBON!!!
15:29
@MaxVernon Thanks, I knew that the effect of switching the column order in the clustered key was depending on the cardinality of ServerID but I did not expect just one value for that column. He did say "replicated from multiple servers" in another comment.
ooo, sql fiddle got an upgrade
it got the "make it work" patch?
looks like it dropped several DB engines and now only has sql server 2014
tons of links across SO just died
also dopped mysql 5.5 and postgres previous to 9.2
easier to support fewer
@bluefeet i heard the screams of millions of links, dieing in unison
Extra version gets added to the dropdown
@AaronBertrand doesn't load a result set for data
Oh yeah you went further than I did
yup
I've got so many broken links now
lakhs of broken links
millions of lakhs
I wonder if that was intentional or even a consideration when Jake rolled out the upgrade.
Hard to complain about a free service doing whatever they want but I'm sure there will be much gnashing of teeth about it
15:47
Well, that's just like your opinion, man
10,000 for SQL Server 2008
Probably a ton for 2012 across both, too
(I forget what number goes in the URL for that, though)
it's too bad that SE doesn't integrate something in their system
@bluefeet yup, would be really useful
Looks like my authentication to sqlfiddle is dead
May just be my browser, but my gmail account doesn't have a G+ profile so after clicking Accept, it gets stuck on the return auth page
Fortunately, I only have 11 fiddles to my name in answers
16:03
@billinkc same issue for me
there's authentication in sqlfiddle?
@Lamak *was :)
ah, right ;-)
16:41
The error messages you have provided are not from the alert log. There is a problem with the database, and we need the information from the alert log to diagnose what the problem is. I have provided a link in a comment above with instructions that tell you how to find the log file. Help us to help you. — Phil 54 secs ago
Like pulling teeth. Only more painful
How are you all? Been very quiet in here over the last few days...
It's because you stole all of our people for some bitter sequel thing
17:02
what?, I didn't get that one
He refers to SQLBits, where many people were in the UK eating fish and chips and drinking warm beer.
I drank no such thing - I abhor the lack of refrigeration. I had Carlsberg (barely beer) and Strongbow (cold, not room temperature, cider).
ah, thanks for the explanation
getting a new hard drive at work = a day of little productivity
I'm getting a new pair of monitors today, so will suffer the same this afternoon
But then my productivity tomorrow, whoa, watch out!
17:14
@AaronBertrand i was wondering if you'd be available for a quick query plan question
Depends on how complex
(Recovering from time shift and catching up on a lot of things on the homefront)
@AaronBertrand no worries then. it's complex, but i'm perplexed by a query that i tried to rewrite with no CTEs ... it has a different plan shape
Can you send me the two .pesession files?
(You have actual, post-execution plans, right? :-))
17:33
@AaronBertrand sorry for the delay. i have multiple queries in one .queryanalysis file
query away
@Aaron is there a way to add 2 time values? (apart from the hack I did here?): dba.stackexchange.com/questions/94787/…
(converting both to datetime and the do d1+d2)
@ypercube no, they're not numbers.
It's a pity that SQL-Server does not have a type for time intervals (and more functions for date and time arithmetic)
17:48
@ypercube that's what clients are for
I don't agree.
18:00
that's fine. i was just being a bit silly
@swasheck how are stats? Are they up to date? Estimates seem way, way, way off. Also if you look at some of the parallel operations, they are only being processed on two threads, and the other 4 do nothing. Is this really a 6-core processor? Or 6 vCPUs assigned to a VM? Is there NUMA? Have you tried with MAXDOP 1 or MAXDOP 2?
Swasheck runs as MAXDERP = infinity
2
^^ This just about covers today
@bluefeet Sounds about right. Bought a Surface Pro with mine. The rest went to the wife and whatever she feels it needs to be spent on.
18:26
@bluefeet are you getting an iWatch?
@Lamak I have a watch problem so I'd really like to get one. But I really need a new computer at home
a watch problem?
I love them. I have a lot of them
ah, that kind of problem
@bluefeet I also love watches, but only have 2
Limited quantities of the Apple Watch Editiion, Prices from $10,000
wow
@bluefeet You have probably some idea how applicable that is to office conversation here today.
18:31
it's a shut up and take my money kind of day
The new macbook is very impressive. But for me, the apple watch is kind of hideous
That new Macbook is insane
yup
JNK
JNK
I didn't follow the announcements
:(
I thought people were joking about the 10k price tag for one of the premium watches
I think I'll concede that must be a real price point with as much chatter I see about it
18:35
Many people thought so
JNK
JNK
that sounds like one of those Onion articles that comes true in a few years
like "Gillette says 'Screw it, we're going to use FIVE blades!'"
People hate on Apple, but I don't see any other company making laptops that even begin to rival that new Macbook...
@JNK thinnest laptop ever (thinner than all airs), 12" screen, 2308x1440 I think
< 2lbs
well, it's "some group of people" hate on Apple and "some other very large group of people" loves Apple
JNK
JNK
@AaronBertrand How many mortgages
18:39
@JNK $1299/$1599
JNK
JNK
oh that's not bad
nope, it's really impressive
I will not buy one of those for me until they get past the 8GB cap. But I will probably replace my wife's 2007 MacBook with it, even though she says "it still works fine."
I want to do things that a man should not do to machines. Drool
18:52
confused much?
Teradata in the title but tagged with MySQL? What database are you using? — bluefeet ♦ 34 secs ago
our thickest pancake yet http://t.co/JPVabVh1qB
@AaronBertrand stats are up-to-date but sample rate is crazy low due to large table ... 6 NUMA-node AMD POS
There are 6 NUMA nodes? How many cores in each node?
(I miss the "power of 2" days.)
Ahoy Diplomat in chief
wtf
This is teradata but I saw it was suggested to tag in MySQL. I think they are very similar. — angelcake 3 mins ago
I think they are very similar
19:01
think is a bit generous there
@angelcake Teradata and MySQL are not similar in any way except they execute sql. — bluefeet ♦ 10 secs ago
Those stupid suggestions need to stop.
@AaronBertrand 4
So try the query with maxdop 4
Or maxdop 1. I can't even see the whole query but that looks like a monster (the join diagram makes me nervous). I'd have a hard time making suggestions at the table / query level.
@AaronBertrand it's awful
@billinkc update @billinkc set [a-hole] = 1
2
19:10
Or just declare @billinkc [a-hole];
@AaronBertrand true
Let's agree that the best option would be DROP TABLE billinkc
@AaronBertrand so run both the CTE and nonCTE with maxdop 4?
nevermind. someone just added an index. hooray for release management
Yeah, worth a try, also worth a try with no parallelism at all. Your big problem is the seek on fund_bmk, if you can't get accurate stats (they're 1,179% off!) you may be better off trying FORCESCAN or create a filtered index that can support the query (you'll have to manage those stats yourself though).
@AaronBertrand thanks. yeah - the hinted query uses forcescan
19:18
Oh yeah look at that - it shows how terribly useless estimated costs are too
JNK
JNK
Has anyone here had to hire a DB professional recently? If so what resources did you use?
@JNK We hired an iSeries DBA, does that count?
JNK
JNK
Close enough
I may be hiring a Sr. Developer/DBA soonish
@JNK I believe we posted the ad to our website, monster and used a recruiter. Do you want me to try and find which one?
JNK
JNK
19:23
And prior attempts have had bad luck identifying good people
@KrisGruttemeyer nah I'm just trying to get a general idea of what works
@JNK 100% remote?
JNK
JNK
I think part of our issue here is we are close to NYC but not IN NYC and finance sucks all the talent to Manhattan
@JNK I think the recruiter was what finally pulled him in. We got incredibly lucky, though. He has 30 years experience, worked for IBM and did consulting for about a decade.
JNK
JNK
@swasheck probably 0% but I can find out :)
specifics are being worked out
im gonna delete that for posterity
you are gonna delete it for posterity?
JNK
JNK
19:27
yes
PURGED
I don't think that word means what you think it means
Posterity hears ya'. Posterity doesn't care.
Damn kids these days
JNK
JNK
@billinkc I want to make sure nobody in the future ever sees it
it's the inverse of preserving for posterity
I think your first mistake was in thinking anyone in the future would seek out your ramblings.
functions galore
0
Q: SQL Statement running extreamly Slow

user3120232Okay I have look through several posts about SQL running slow and I didn't see anything similar to this, so my apologies if I missed one. I was asked about a month ago to create a view that would allow the user to report what hasn't been billed yet, and the SQL is joining 4 tables together. One i...

JNK
JNK
19:32
@billinkc It won't be my last.
@bluefeet Also can you WHERE SUM()? Should be HAVING no?
I long for a GetBillInKCProgress method to be standard everywhere — billinkc 16 secs ago
And deleted for prosperity
JNK
JNK
@billinkc Delete it to 10 friends and money will come your way this month.
I don't have 10 friends
Even in binary
JNK
JNK
@billinkc Not even "friends" on facebook?
You have at least 10 spambots following you on tweeter
And I paid good money for them
19:37
@AaronBertrand 3310 on SO and 101 on DBA.SE. They aren't broken actually but they will now execute in SS2014.
19:48
25 or 6 to 4
@AndriyM 9987 and 206 for mysql 5.5
0
Q: SQL Date Conversion Query

tia97TimeStamp is a numeric datatype field in my table R for example: 2445302102010 (02/10/2010) I am trying to query by a date range of 02/09/15 - 02/15/15, the problem is I am getting results from 2010? Select distinct R.CID, RIGHT(R.TimeStamp,8) from TableRev R...

I don't always store dates as numeric, but when i do....
@ypercube It's amazing that the figures for SS2008 are exactly the same. To the point that one might think there's some kind of mistake :)
JNK
JNK
@KrisGruttemeyer no as STRINGS
because they want to keep those padded 0s in there
because datatypes are for suckers, man!
@JNK It's just sad, really. Just use the correct datatypes. You can convert, cast and reformat to your little heart's content, just make sure the damn underlying type is right.
/rant
Just drives me crazy when people use a spatula to rebuild a car engine.
JNK
JNK
20:01
"But I'm really comfortable with spatulas!"
HA!
It's not even CCYYMMDD
@AndriyM Or I did a mistake. I thought !3 was for 5.5
It's all bass ackwards
0
Q: SQL Server hitting System.OutOfMemoryException but server has 10GB free

jcollumSQL Server 2012 11.0.5058.0 I'm running a very large batch of queries (about 2700) to simulate a user doing X, Y and Z in our app. I have SQL Server's max memory set to 2147483647 MB. However when I run the query I get several messages like so: An error occurred while executing batch. Error me...

Answer: Don't have SSMS return 18 million rows to the results window
@ypercube 19279 and 298 for MySQL 5.5
And the award for the most obnoxious comment goes to
Help me understand: your source query has some identifier to indicate what catalog the data resides in. These catalogs have consistent entities in them and you need to run some query across them all (perhaps a billing report). Is that about accurate? That sounds accurate. As Ako asked, is it possible that Client20 suddenly accumulates so much data you need to push them to their own server? I am not sure to be honest, It's a pretty simple lookup (at least with script component) and the results are saved as flatfiles if that info is of any use — August 5 hours ago
20:16
it wasn't what I was expecting
I should delete my comment to make that look really insane
actually, you should delete your comments, period
This query works great except I have encounter two issues. 1. I get an error "Athen — angelcake 45 secs ago
How can it work great and they get an error? ...
maybe before he got many errors
many errors. work great. so much broke.
JNK
JNK
20:35
@billinkc And the award for the most obnoxious comment goes to I'm not sure what you find obnoxious about it
It's like the people who respond inline to emails - see my response in chartreuse below
JNK
JNK
we have some of that here actually
I've done it if there's a bullet list of Qs in the email or something
@JNK the OP copied all of Bill's comment and then added the answers at the end
That I can at least understand. I have people that weave sentences into existing paragraphs
JNK
JNK
@AaronBertrand the OP copied all of Bill's comment and then added the answers at the end Yeah I saw :)
5
20:37
@JNK And the award for the most obnoxious comment goes to I'm not sure what you find obnoxious about it Here, allow me to illustrate
JNK
JNK
@AaronBertrand note that my response to him was in the same format
Sorry, jetlag
JNK
JNK
Its' fine
How was Britain
Just when I'm finally giving up on being distracted and doing some work, y'all go and entertain me
JNK
JNK
I saw you and Kevin Ghostbusting
20:39
It was fun. Love that conference. Cool folks.
55,892,200 rows = 1.25% default sample rate (SQL Server 2014 Compat Level 12 TF 2371 enabled) ...
JNK
JNK
@AaronBertrand That seems like one of the better conferences
It also saddens me I couldn't go and probably won't make PASS this year
JNK
JNK
yeah I know
We spent a lot of new money on stuff recently
like monitoring tools from this company based in NC
I'm going this year, my first one.
20:46
Then I meant YAAAAAYYY!
JNK
JNK
lol
We also dropped some money on a SSRS config manager thingy too
so database funds got used up
Here's a Ferrari. Nope, you can't have training manuals until next year
@JNK Mine didn't even make it into our budget. Guy who was in charge of it (higher up, after I asked repeatedly if he wanted me to do it) 'forgot' to add it to the budget. Might not be able to get our prod server off of 2005 til next year
@JNK Company in NC that does SQL monitoring? Has to be based out of Raleigh or Charlotte
JNK
JNK
@KrisGruttemeyer It's SQL Sentry, was messing w/Aaron (who works for them)
which BTW I highly recommend
@JNK Ahhh, I see now. Carry on
20:50
$1495 is the current registration price, can usually find someone offering 150 off of that price. Then it's a question of airfare, hotel and vacation
@JNK Yeah, we have foglight and redgate. I've tried SQL Sentry but can't convince the higher ups to buy it.
Alright, time to head out on this scorching 43 degree day. Nice to see that white crap on the ground start to melt. Have a good one guys
JNK
JNK
Good night
21:33
@KrisGruttemeyer I hope you don't have *= and =* in your code, like this guy
0
Q: Moving from SQL 2005 to 2012

VictorHere is my Stored procedure written in SQLServer 2005. Now we are moving to SQL2012 and getting bunch of errors where I used *=. I know I have to use LEFT OUTER JOIN syntax, but somehow I am stumped. Can someone please help me ? Also getting error on where I used ABS in where clause. These 2 li...

Wasn't the syntax deprecated from 2005 version? or it was in the 2008 version?
21:58
@ypercube deprecated in 2005, but last working compat mode is 80. So can't be made to work in 2012.
@AaronBertrand So this guy must have had the db in 80 mode?
@ypercube yes, that's right. I can validate later but not at the moment.
YOU SUNUVA MOTHERLESS GOAT! I FART IN YOUR GENERAL DIRECTION
@AaronBertrand Thnx. Just curious I was.
@AaronBertrand Perhaps you should make a blog post about how to convert such queries. I'm sure there are still more dbs out there and it's going to get more difficult, when they finally decide to upgrade to 2012 or 2014.
I searched the Net (to confirm that *= is LEFT) and it wasn't easy.
22:18
@swasheck I fart in your general direction!

« first day (1525 days earlier)      last day (3642 days later) »