« first day (1301 days earlier)   

12:02 AM
Hm. Have a table with two varbinary(max) columns, there's no non-table-scanning way to break down the average (or overall) size of each column is there? (No indexes on either column, since A: it's binary and B: the total size of the data in that table is over half a TB)
 
 
1 hour later…
1:24 AM
Don't know if I've posted this here before
 
 
3 hours later…
4:53 AM
-1
Q: SSIS Environment variables SSMS vs the OS

user2516569As I'm cramming for the 70-463 exam on SSIS, I'm making my way through the mess of terms they've introduced. Now at the project/package deployment. Can anyone explain me the difference (if there is such) between: the "environment variable" which is defined with the OS (control panel -> system -...

 
 
4 hours later…
9:02 AM
I forgot about this and hit it again today. At least it reminded me to accept your answer, thanks :) — Jack Douglas ♦ 52 secs ago
 
9:46 AM
@MarkStorey-Smith MS licensing can best be described as "Byzantine", so it's really anyone's guess
 
@SimonRigharts Well, until I read that post I was 100% sure this wasn't allowed. Paul Randal seems to think the same (albeit back in 2009). Maybe there's been a change in the 2014 terms?
Has anybody encountered ActifIO in a SQL Server context?
 
10:13 AM
After 3 days and 4 answers, they decided to change the code. I don't like when they do this: dba.stackexchange.com/posts/72367/revisions
Some of the answers as are relevant, despite the change, but still ...
 
11:10 AM
If I want to feed multiple files to PG's COPY statement, is something like this the best I can do?
2
A: Copy multiple CSV files into postgres

Clodoaldo NetoIn linux pipe the output of the files listing to psql. Make copy use the standard input: cat /path_to/ys*.csv | psql -c 'COPY product(title, department) from stdin CSV HEADER' Look for the equivalent in other OSs

COPY itself does not seem to support multiple files. Or should I pre-concatenate the files?
I.e. is there any disadvantage in reading from standard input?
 
11:42 AM
@ypercube Not very useful response back from the estate agent mob was just that order of preference for areas is Crystal Palace, Norbury, Thornton Heath. No suggestions as to where in those areas is good/bad unfortunately.
 
@MarkStorey-Smith No worries.
We have wondered around all day on Saturday with my wife and saw several houses/flats.
 
11:54 AM
@MarkStorey-Smith We lean towards to not moving there but staying in Woking.
 
JNK
12:44 PM
The UK has great place names
 
0
Q: Replication saying my snapshot Publication Does Not Exist. But it does exist

dkimbrellI used the SQL Server (2008) wizard to create a new snapshot publication on my sever called "BlinkViewsOnly" in the BlinkDataMaster database. Publication was created successfully. I then used the wizard to create 1 subscription to my laptop. Results messages all appear to be successful, and I ...

VtC (off-topic)
 
1:21 PM
@JNK do we want the question above?
 
@JNK Certainly colourful.
And mostly old too.
 
JNK
@Kermit seems to be in our wheelhouse
 
@Jack, @dezso or any other Postgres user around?
Is it normal I see "Running clusters: 9.1/main" when I check the Postgres service?
 
1:43 PM
@ypercube I used to live in Woking. It's OK. The nice areas are Horsell and St. Johns. The train links into London are pretty quick as well. If you hunt for digs around the brewery road area you can get a nice house that's walking distance to the train station as well.
@JNK I used to live just up the road from here:
 
@ConcernedOfTunbridgeWells Knockhot and bottom? Nice
 
@ypercube Pratt's Bottom no less.
 
@ConcernedOfTunbridgeWells ha, my last name is Pratt :)
 
@bluefeet If you ever come to London I'll take you to visit Pratt's Bottom. Not that there's anything much there.
 
1:59 PM
@ypercube where do you see this?
 
@ConcernedOfTunbridgeWells I'll keep that in mind.
 
2:40 PM
Cdr Sarah West must really love her seamen
2
 
@swasheck That really needs to be followed by fnarr fnarr
 
fair enough
 
A coinage from Viz made it into the OED.
@swasheck It this case it really does help to be British and of a certain age, although Viz made it as far as New Zealand. Never seen any reference to it gaining any traction across the pond.
 
i'll just believe you that it's frickin' hilarious ;)
 
2:56 PM
@ConcernedOfTunbridgeWells tl;dr but would it be a more explicit version of Mad magazine?
Assuming you have a frame of reference for Mad...
 
@billinkc Sort of, with a very North-England cultural basis.
Really big on puerile toilet humour and satire on popular culture.
Very un-PC and occasionally hysterically funny.
 
is this GROUP BY valid in oracle?
0
A: Row that should have a 0 count not showing

LingerHow about joining to the tables instead of using the WHERE clause: SELECT s.Section_Id, s.Location, COUNT(e.Section_Id) AS Enrolled FROM Course c LEFT JOIN Section s ON c.Course_No = s.Course_No LEFT JOIN Enrollment e ON s.Section_Id = e.Section_Id WHERE c.Description = 'Project Management' ...

 
3:12 PM
hmm, I wonder why?
0
Q: Bad performance on sql query

Jean Carlo MachadoSELECT orcamentos.id FROM usuarios, clientes, empresas, orcamentos WHERE ( ( clientes.nome LIKE '%VARIAVEL%' AND clientes.email=usuarios.usuario ) AND orcamentos.cliente=usuarios.id ) OR ( ( ( empresas.razao_social LIKE '%VARIAVEL%' ...

 
@Kermit which bit are you worried about?
 
@Kermit I'd expect it to be. It groups by three columns, of which there are only two in SELECT, but that is fine in SQL. You can't select more (non-aggregated) columns than you group by, but you can group by more columns than you select.
 
@AndriyM (and @JackDouglas) okay that makes more sense... c.Course_No didn't appear in the SELECT clause.. i don't think that would be valid in sql server
 
@ypercube how are you checking it?
 
@dezso @JackDouglas running: service postgresql status
 
3:17 PM
root@dallas:~# service postgresql status
9.3/main (port 5432): online
I guess it's fine
 
@JackDouglas Wrong Fiddle, it seems
 
@JackDouglas I get the same as yours in my Ubuntu desktop but in Debian:
> root@dbserver:~# service postgresql status
Running clusters: 9.1/main
 
mine's a vanilla Debian Wheezy install with 9.3 installed on top
 
@JackDouglas makes more sense
 
3:21 PM
@Kermit It's valid in SQL-Server as well.
 
well bob's your uncle
0
Q: Unable to print SQL Server 2014 Reporting Services in IE8

Scott SledGimp GravesMy company has upgraded all of its client databases from sQL Server 2008 to 2014 and we have now found that none of our clients are able to print the reporting services reports in IE8. Unfortunately we are required to support the legacy app but I have been unable to find a resolution. When clicki...

 
@billinkc was you comment directed at me?
 
Nope
 
ok. just couldnt tell. i wouldnt care if it was, i was just trying to decipher the meaning
 
4:07 PM
@swasheck Mine was.
 
cannot see it
 
That's because it was never really there.
 
keyser soze of comments
 
The worst part of my job is after I've submitted my code being forced to go back and tag a little WITH(NOLOCK) after ever FROM or JOIN.
With each add I die a little inside.
 
or you could ...
SET TRANSACTION ISOLATION LEVEL READ UNCOMMITTED at the beginning
 
@billinkc can I (ab)use your SSIS knowledge, please?
 
sure
 
is there an easy way to check if a package is using Enterprise features?
or I have to parse its XML content? like here
 
If by enterprise features, you're just looking to see if they used things like Fuzzy Lookup, etc and not digging into a SQL Task to see if they turned on compression or something, you could grep the file looking for the component ids
 
only for SSIS features, not for the SQL Server itself
 
4:27 PM
There's very few EE features in SSIS, personally never had cause to use any of them. It'll be what, the Fuzzy* components, CDC if you're in 2012 and ... I'd have to look at the documentation
 
the instances will be mostly 2005 and 2008R2 and we've got packages developed by other devs, so I've no idea what's inside :)
 
2008 ComponentClassIDs
Fuzzy Grouping: {0A1BDBA0-9C6E-4AD7-97B1-E0E79E462629}
Fuzzy Lookup: {5056651F-F227-4978-94DF-53CDF9E8CCB6}
2005 {0A1BDBA0-9C6E-4AD7-97B1-E0E79E462629} / {5056651F-F227-4978-94DF-53CDF9E8CCB6}
Unless Zane's aware of other ways to find it
Looks like Data Mining and cache connection managers (surprised me) are also on the EE feature list msdn.microsoft.com/en-us/library/cc645993(v=sql.105).aspx
 
alright, thanks. I was thinking if maybe there's a bit IsEnterprise somewhere and I'm just missing it.
 
You made me snort
 
hope is was related to a smile, too :)
 
4:42 PM
 
@billinkc when the day is done and you want to run ...
 
5:08 PM
@swasheck ahhh but I can't
 
por que no?
 
@billinkc nope.
 
0
Q: Select returns inconsistant data

lyrixxHow is this possible? insight=> select id from analysis_analysis where status != 'finished' and status != 'archived' and begin_at < '2014-07-28 17:23:27' limit 1; id ------- 46632 (1 row) insight=> select id from analysis_analysis where id = 46632 ; id ---- (0 rows) I execute theses q...

those darn bugs
 
Love when I get the 'I desperately need you help this doesn't work. Let me pick a decidedly inconvenient time to work through it with you." Requests.
Starting to write a presentation on Reading execution plans for beginners and it's trickier than I thought. I'm trying to decide where to go and how deep to dive.
 
5:35 PM
The feedback I received was that I wrote my presentation for the alpha nerds because that's who I hang out with. Instead, I needed to simplify it a few orders of magnitude to reach my audience members
YMMV
 
If it goes well maybe I'll submit it to SQLSat MSP as well.
 
@billinkc alpha nerds?
 
Speakers and other a-type, overachievers
 
@billinkc and how do they let you hang out with them?
like, an ego boost?
 
Pity
 
5:57 PM
I was looking at the list for MSP is looking more and more intimidating. However there isn't much for SSIS so there may be a chance.
 
6:57 PM
@billinkc didn't you say you used to use HierarchyID for a project?
 
I did
 
A coworker is telling me there is no way to update it it with a set based operation. I'm assuming this is not true but looking to confirm quickly.
 
Unless I'm misunderstanding their statement, I'd say there's no way what they're saying is correct
 
He's using a loop at the moment.
 
Looking at my code, I only generated a hierarchy id, never updated it but I don't recall seeing anything in the documentation as such
Logically, it might be hard for them to calculate the new value but syntactically, I don't think there's anything stopping them.
My calculation is recursive so it's got that going for it...
 
7:13 PM
See history if you want to see it.
 
dont tell me what to do
 
I already did. What are you going to do about it.
 
:)
 
7:16 PM
It's a double loop
Why?!
@swasheck Point taken... :)
 
This HierarchyID seems like a needless ass pain.
 
@billinkc ssis question for you
 
It seems like just having a relational table would solve the problem with less jacking off.
 
we've got an Excel file that we load in. The Source uses Sheet$A3:O - if I add new columns to the spreadsheet - how can I easily tell it to start using the new columns. I've only been able to edit the XML to include the new ones. Is there some other way I'm missing?
 
7:44 PM
Excel, SSIS and easy never go in the same sentence unless the sentence also contains "can't make it Fing"
Rant aside, you could look at querying the worksheet instead of table access mode
You could also try named ranges or whatever they're called in Excel.
SELECT * FROM Worksheet$MyRange (I'
'd have to bingle the exact syntax). Range doesn't have $ in it, IIRC correctly
Also, script approach via 10lb sledge
2
Q: script task in SSIS to import excel spreadsheet

Tim VavraI have reviewed the questions that may have had my answer and unfortunately they don't seem to apply. Here is my situation. I have to import worksheets from my client. In columns A, C, D, and AA the client has the information I need. The balance of the columns have what to me is worthless inf...

 
ok. I'll take a look but that'll involve more changes at this point.
The XML edit was quick
I just wasn't sure if I was missing some easy fix
 
Negatory. My feedback for SSIS pain points to the MS folks was "Excel in general. I don't even touch the SSIS questions that pertain to using Excel as a source or destination."
 
It's surprising how poorly they play togather.
 
good to know. I'll stick with my XML edit until later. I don't feel like completely redoing it right now
@Zane it is microsoft
 
Excel doesn't play well with a lot of things from what I can tell.
 
7:52 PM
Excel is a great tool for data analysis. But not for data storage/ETL
 
This hierarchy id crap is the worst project. I can't believe how poorly this is all set up.
 
The problem with Excel is that they haven't been able to give it a clean rewrite due to a need to retain backwards compatibility. I have to believe that if they could take all the lessons learned and just start with a not-stupid file format, world peace would be achieved
4
 
@billinkc but then the spreadsheet I made back in 94 wouldn't be compatible!!!!
 
@MikeFal yo
 
troof
 
7:53 PM
@swasheck oy
 
what's the frequency, kenneth?
 
@swasheck REM blows.
 
8:15 PM
@MikeFal ... so ... how're things
how're the talent-poaching ventures?
 
@swasheck Things are mediocre. Stressful weekend. Poaching.....I need to get an update.
 
sorry man. work-related stress?
gosh. why is my fullscan stats update taking so long in this 16bn-row table
 
@swasheck Regrettably, no. But it will be fine.
 
well, sorry nonetheless.
 
JNK
@swasheck did you try NOLOCK?
 
8:22 PM
@JNK maybe this is what your histogram should be. maybe not, though.
 
JNK
@swasheck "This is closeish!"
 
and i seem to recall that @PaulWhite said something along the lines of "a good estimate is acceptable" in a blog post. ergo, UPDATE STATS WITH NOLOCK!
sadly, all my stats updates show as running with READ COMMITTED
 
JNK
seems legit
 
This loop crashed SSMS.
Oh how I hate this issue.
 
loops = You spin me right round, baby, right round.
 
8:27 PM
@swasheck Yes, in my most recent one
 
;)
@PaulWhite i think that this thing can be truly awesome, sys.dm_exec_query_profiles, but am struggling to find a way to actually use it properly.
i've read some blog posts
it seems like i would have to be excellent with timing things in order to capture final output
 
@swasheck Yes it needs tool support to make it easily usable/informative
 
JNK
@Zane seems to be last page in a tutorial to be fair
 
8:44 PM
@JNK Just seems odd that it essentially a link to the first page.
 
@Zane Oooh, are you playing with hierarchyID?
We have some of that on this project that Bill was working on.
 
@mmarie No I'm trying to give emergency surgery to a project that is populating HierarchyID in the worst possible way.
 
@Zane Eew, that sounds unpleasant
 
<-- goes down the HierarchyID rabbit hole.
 
9:05 PM
@MikeFal NSFW
 
@swasheck That's right, man. Hardcore, full-frontal scripting.
3
 
9:21 PM
@MaxVernon let me know if you find any gems in that hole as all I'm finding is misery.
 
^^^ bro. do you even english?
 
9:43 PM
I've been reading and reading and reading but I don't see the value of this datatype. Am I missing something here. It seems just having a table to track the management hierarchy would be infinitely easier to work with.
 
Holy War material right there.
 
and you guys think the sql folks are bad on SO
-2
Q: A bunch of questions about generator-cg-angular

Dan CancroMy app generator comparison has a bunch of unknowns about generator-cg-angular. The True/false items are benefits or attributes that a generator may have or not have. These things might also be 'optional' or not applicable due to other factors in the generator. For the Explain items, I'd like ...

a bunch of questions = 73 in a single post
 
At least you now have a definitive example of "Too Broad" :)
 
@Zane its a varbinary -> not sure how wide it is. I think it is officially a varbinary(max), but I wouldn't be surprised if I was wrong!
 
@PaulWhite that is an understatement :)
 
9:54 PM
@MaxVernon "The type internally holds a binary value as large as 892 bytes that represents the path of ancestors leading to the current node."
 
it's so funny how many people misspell rogue as rouge. totally change the meaning and it's HILARIOUS
 
@PaulWhite over the hierarchyID datatype?
 
@Zane Yes.
@Zane @MaxVernon It's actually a system CLR type. It does use a proprietary binary format for serialization, but that's not the same as saying it is a varbinary.
@swasheck Could certainly lead to some rogue faces
 
yep
we have a few queries that have "gone rouge"
reminds me of this
 
The binary format of the hierarchyid CLR type can be seen in the source here
Actually that's not quite the whole thing, but the ideas are there
 
10:23 PM
@PaulWhite Do you stand on any particular side of that holy war?
 
@Zane All the alternatives have pros and cons. It depends on the project requirements. So, sitting on the fence, yes. I have never had the need to work with hierarchical data in a real system.
 
I feel like just having a table structure to handle hierarchy feels more intuitive to me.
This may have to do with the scenario that I'm working with hard to tell as this is my first time using HierarchyID
 
10:43 PM
Simplest is often best.
 

« first day (1301 days earlier)