« first day (463 days earlier)      last day (2015 days later) » 

1:36 AM
@NickChammas yeah, there's an issue on SO. Working on getting it resolved (if it wasn't already)
5 hours later…
6:08 AM
@NickChammas It's here
Q: What's the difference between a temp table and table variable in SQL Server?

Eric Z BeardIn SQL Server 2005, we can create temp tables one of two ways: declare @tmp table (Col1 int, Col2 int); or create table #tmp (Col1 int, Col2 int); What are the differences between the two? I have read conflicting opinions on whether @tmp still uses tempdb, or if everything happens in memor...

1 hour later…
7:19 AM
Morning all
8:23 AM
1 hour later…
9:24 AM
@ConcernedOfTunbridgeWells Sorry for the late answer -- I live in Portugal
Morning chaps
Hi Mark
9:54 AM
10:15 AM
@ConcernedOfTunbridgeWells Ta
I don't think so... job running for 10 hours.. ouch
@Phil Crap infrastructure is the bane of my existence. Why, why, why can't they just provision a £20,000 server that's actually fit for purpose.
Doesn't affect DB licensing (already got that), doesn't affect operational costs (most of which is running the database). Uses more power than a SAN? srsly?
beatings, beatings, beatings.
10 hours may be 100km with a well trained runner. If the 2 servers are not in different continents, a hired runner or a pigeon to carry a USB stick with the (updates) file can be solution, can it not?
Obviously I haven't woken up right today. I feel an urge for downvotes and nasty comments ...
1 hour later…
Yeah, I think I've seen it in a SO answer
Q: How do I upload huge files across the internet without using P2P?

Brien MaloneI work remotely and have 44GB of media files that I need to send back to my office. There are lots of free services out there that can handle up to 2GB, but I haven't seen talk of anything larger. We both have 50mbps+ connections, so I would rather not mail physical media (though, that is an op...

Probably several
12:02 PM
Good morning, fellow data-ers
Not really impressed with that answer to the question i linked in here before
1 hour later…
1:08 PM
@Phil neither was I
1 hour later…
2:19 PM
@Phil That's because the best answer, with the furthest-reaching benefits to the question is yours.
Was that answer converted into a comment?
OP replied and said that it wasnt an option
@ypercube yes - I did that. right move?
I must be in a different question
i'm talking about this one: Improve performance for my plsql block
@jack: Well, it was not really a good answer. I don't know how @Rainier sees it though.
2:31 PM
well, it also helps if i practice some reading comprehension
you said move answer to comment (not vice versa)
Get them to sort out the firewall timeouts then. I'm not a big fan of writing nasty workaround SQL (and increasing remote DB roundtrips as a result) because of infrastructure issues. Sorry to be blunt! — Phil 5 hours ago
I wouldn't like my answer ot be deleted (or turned into a comment) - if I thought it was an answer :)
@ypercube Do you know @Rainier?
Alex said "The infrastructure thing is i think i'm affraid is out of the question for this matter :) ill have to do the best with what i have :) sad but true" in a comment - therefore Rainier's answer was "Not an answer" in my book
not his fault though
I'm not sure there will be a good answer for that
2:36 PM
In such a handcuffed situation, it should probably be collaboratively escalated to a decision-maker a few levels higher
@Phil if that turns out to be the case then we may need to consider closing the question
@Phil i like that @Erkan called you Philip.
I may upvote him just for that
I'll wait for the explain plan output before i post more on that one
@Jack: I agree. (don't take me too seriously today...)
i suspect it's a hideous remote plan
2:58 PM
@ypercube why isn't your comment on this question an answer? :D
I am so poor at indexing strategy. I know it's supposed to be basics, but I'm seriously like o_O even when I read simple things like @ypercube's comment
basically the 'always' field only has two possible options (true or false). when choosing an index, you want to be discard as many entries as possible. I guess MySQL thought that if you can't discard 95% of the rows, it's not worth it :)
i slowly deduced that, but would have never pulled that out as an answer/comment
tbh, i would think being able to discard 50% of a 30 million row table would be worth it to only have to check 15million rows.
not saying this guy's case is 30-million strong, just in general
@swasheck But, just to add to the complicated nature of indexing strategies, Using a Low-Selectivity BIT Column First Can Be the Best Strategy
3:13 PM
@MarkStoreySmith This is where it throws me. I need hard-and-fast rules. Read through use-the-index-luke.com and thought that I had a good idea and then thhhpppptttttt something changes and makes a large exception. Something along the lines of "I before E except after C and sounding in A as in Neighbor and Weigh ... and also weird because weird is weird."
I figured I would join and defend my honor
That is not my query!!! :-)
@josh no one said it was :)
@DTest that was helpful, but unfortunetly changing the index didn't help. I think his entire database design needs to be rethought
@DTest OK just making sure :-)
This developer has a tendency to just add random columns to databases without rhyme nor reason
@JoshGitlin on-the-fly denormalization
@swasheck As much an art form as it is a science unfortunately :)
3:16 PM
@JoshGitlin Good reason to not let developers have access to production db's
@MarkStoreySmith Clearly. That's ok, I just need to dedicate the time to really sitting down and conceptualizing it. Once I do that then I should be able to work my way through it.
One of the reasons why SQL is such a pain for cargo-cult developers
@DTest Sadly this developer is also CEO of the company, LOL
Thankfully not your company, though ;)
@JoshGitlin So then he should understand the explicit risks of such activities, no?
3:17 PM
@Josh Nice :)
@TimStone we hope
@swasheck He will when he gets my bill and his customers complaining that the server is slow :-)
Ego overcomes common sense
What's good today fellas?
True that
3:19 PM
A friend told me that in his first position, his responsibility was dealing and cleaning a lrage amount of MS-Access databases. All had been created by the previous No#2 (CEO?). All tables had names: George1, George2, ... GeorgeN and all columns were column1, column2, ... (guess who was named George :)
I know that my sinus headache rapidly turning into a sinus migraine is NOT good
@jcolebrand educating me in the art and science of indexing, while tossing a few jabs at CEOs who want to code
@TimStone your Skype answer was perfect, don't be shy, post!
I'm afraid of all of these real DBAs ;)
3:22 PM
i'm not a real dba, but i did stay at a holiday inn express
Crikey, I'm grumpy. Spent all day writing reports :/
@JoshGitlin I gave it a go without testing (yay me). The idea is to break it into different queries that on their own would use an index.
@DTest I guess my version of MySQL is too old?
> #1064 - You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'GROUP BY movies.id ORDER BY listorder' at line 11
naw, probably bad syntax
This works great, from @TimStone:
3:31 PM
@Josh: First, you should get a clear answer on what the query is supposed to return. As it is, it has undefined behaviour/result-set..
(SELECT movies.* FROM movies JOIN showtimes ON movies.id = showtimes.movie WHERE showdate >= '2012-4-11' AND showdate < '2012-04-18') UNION
(SELECT * FROM movies WHERE always = 'true')
) AS movies
WHERE site = 5
GROUP BY id ORDER BY listorder
but he's being shy :-)
@ypercube Yeah, I know. Dude won't return my calls, LOL
meanwhile his server is SLOW
meeting, bbs
@JoshGitlin yup, derived table is the way to go. the only thing I changed in his is to use site=5 on each subquery. especially in the 'always=true' query. Am hoping site=5 is a better selective index :)
I updated my answer with that. My derived table is aliased different than the movies table to avoid confusion as well.
4:17 PM
Oh god why
So. Moronic ERP software fixes a bug where they made a comment field too short in the database
Instead of expanding the column for comments, they add more rows of the same size, so you get more rows with longer comments
Which makes outputting the comments an unholy nightmare
@BenBrocka clearly a business case for MongoDB or Cassandra. this is why sql sucks
4:59 PM
Would you guys be interested in a question like this?
Q: Database Table Prefixes

DoctorMickWe're having a few discussions at work around the naming of our database tables. We're working on a large application with approx 100 database tables (ok, so it isn't that large), most of which can be categorized in to different functional area, and we're trying to work out the best way of naming...

hmm, seems to be a very opinion-oriented question, which imo doesn't bring anything to dba. Just my 2 cents from a non-oracle guy.
Ok, thanks. Didn't know if I should recommend moving/asking the question here or not
@Rachel his requirements seem a little too vague: "It's possible (although not guaranteed) we'll have a number of different applications which could require the same data structure, or even the same data, from one or more functional areas but not all of them."
@Rachel Oracle deities like @JackDouglas may be able to speak to this better, but it seems like you may have a good answer in there. It is an architecture decision ... I defer to @JackDouglas or @jcolebrand
@swasheck I prefer the term "Oracle Bod" or "Oracle Monkey" :-)
we'd probably close NARQ if it arrives here IMO
5:13 PM
@JackDouglas Bod? Not like BOD (Board of Directors), right?
@swasheck UK slang perhaps?
@JackDouglas Agreed in the current form I'd vtc it :)
@JackDouglas Bod in the form of "body" is familiar to me but I was just trying to understand the connection with Oracle. :) Monkey just sounds so insulting, from an evolutionary perspective.
@swasheck insulting to whom?
5:16 PM
@JackDouglas the addressee. i've only heard it as somewhat (though not entirely) derogatory. "Oh. That's a job for the net monkeys."
hah, it's amazing to me how information comes. I just learned earlier today that bitmap indexes are decent at low-selectivity indexes...Then I read in this data-warehouse book that bitmap indexes are one reason to denormalize dimension tables.
@swasheck ahhh - I thought you meant insulting to the monkeys :-)
@DTest bitmap indexes are evil
at least they are on Oracle
I would hesitate 20 times before using them in MySQL :)
at least
back to the old days of page-level locking
fine if you are not doing any updates of course
right, i did mention it was in a date-warehouse book, right? At least I assume those are read-centric
5:36 PM
Dtest: are there bitmap indexes in MySQL?
@JackDouglas Bitmap indexes are very expensive to update; they're really designed for star schemas where you're loading in batches. You can drop or disable the indexes for the partition, load data and then rebuild the indexes. Not much use outside of data warehouse systems but quite good for certain types of queries on star schemas.
@ConcernedOfTunbridgeWells agreed
@DTest was "I just learned earlier today that bitmap indexes are decent at low-selectivity indexes" also from the DW book?
doo be doo. Ignore this entire conversation please. BITMAP index <> index on BIT column. Off to hide in a corner

« first day (463 days earlier)      last day (2015 days later) »