| |
| |
| |
|
Page: 1 2 3 4 5
Comments:
<0> ok got a question
<0> http://sql-servers.com/nopaste/?show=331
<0> view code is there
<0> this table in the view SOP030300
<0> has like 6 mil rows
<0> select sum(PRICE) From vSalesOrderItems
<0> this query takes like 5 minutes
<0> how could i make it faster
<0> i tried putting an index on the 4 cols in the view, but doesnt help
<0> doesnt seem to be using it
<0> am i invisible?
<1> yeap
<1> or you just ****, im not sure
<0> well help me out either way
<1> I dont know mssql
<2> damn I need to add some indexes here :-\
<0> why would this view not be using the index i created on the same 4 columns?
<2> how do you know it's not?
<2> 6M rows takes a while to process
<0> cause im looking at the execution plan
<0> all it shows is a table scan of the main table
<3> anybodyhere us mse sql 2k5? trying to figure out how to set the compatability level . getting an error specified @subsystem_name SSIS does not exist when trying to copy a database from a ms sql 2k server to a mssql 2k5
<4> 0y
<4> ok...
<4> so anyone know about coalescing records?
<4> specifically i've got (Record 1: [1]
<4> er..
<5> what is a "coalescing record" ?
<4> specifically i've got (Record 1: [1][2][NULL][6]) and (Record 2: [1][2][9][NULL])... need to combine into [1][1][9][6]
<4> can't do distinct cuz neither one will catch
<4> could do @var=SELECT TOP 1 for each... but for a lotta records, it gets ugly (technically the last two fields are data, and the first two fields are FK identifiers... so they'll be joining against another dataset)...
<5> use min/max?
<4> ?
<1> why can you use sum?
<4> hm
<4> another great suggestion!
<4> dude
<4> that's perfect
<4> SUM for the FK identifiers
<4> hm
<6> CoderMan, it may actually be more cost-effective to scan; it depends on many factors
<4> thx coj... it'll take some playing, but that could work quite well
<4> lvk: what's your idea w/ min/max ?
<1> yay! I did work today
<2> how are you going to put that together, to apply MIN to the SELECT and then stuff them together?
<3> does SSIS exist for sql 2k ?
<4> does min() catch NULLs or not?
<5> same as with SUM just use MIN or MAX instead
<6> no
<4> hm
<3> does SSIS exist for ms sql 2k ?
<6> DTS
<4> yea might use min/max instead just to be sure i don't accidentally combine data... but the general idea from you two is a *LOT* closer than i was :)
<4> thx
<3> would it be called SSIS in MS SQL ? or is that new terminology for sql 2k5?
<1> min/max wont return null unless all rows are null I think
<3> cause i'm getting the message the subsystem "SSIS" does not exist.
<4> coj: yea that's sort amy thought
<4> Arrakis: new to 2k5
<4> Arrakis: before 2k5 the solution was DTS
<3> ok
<6> Arrakis, you can run DTS packages on 2005, but you cannot run SSIS packages on 2000 :(
<4> true
<4> DTS would be backwards compatible :)
<3> yeah :S TheGamble - i'm trying to do a "database" copy
<4> or rather... forwards compatible, i guess
<3> i copied the tables via export data
<6> heh; cruel joke by Microsoft
<3> but it wouldnt let me choose to copy FK / default values
<3> so i have 20+ tables that all fail are are no longer related because this **** didnt get transferred over
<6> yes, 2000/2005 is not a painless transition
<3> I am finding that out :(
<3> i wonder if maybe i need to have sql 2k push it to sql 2k5
<3> instead of have sql 2k5 pull from sql 2k
<4> hehe
<4> prolly a *ton* easier
<2> aahh push it / PUSH IT REAL GOOD
<4> though there are some new capabilities that you might be expecting
<1> heh
<4> it's a question of "requirement" or whatever for the app
<3> well the web server has sql 2k5 and .net 2.0
<3> so i gotta pull all of this crap out of 2k and into 2k5
<3> which for the most part worked after 24 hours of gruelling shoveling
<3> but i find out that all referential - all the default values - constraints and **** - didnt get tnrasferred over
<3> i wonder if its possible to just script the alter table ****
<3> and apply that
<6> We are currently reviewing the whitepapers and developing transition plans
<2> now I have that Salt-n-Pepa song stuck in my head
<6> hahaa
<2> Can't you hear the music's pumpin' hard like I wish you would?
<2> Now push it
<2> and damn it these tables need indexes! :-\
<3> hmm
<3> looks like i can't
<6> haha!!!
<2> stupid ***backwards ERP retards
<3> i would have hoped sql 2k
<3> cvould have at least conected to it
<3> i mean
<3> hugg
<6> I have a feeling that I have not felt all of the pain yet. One database in particular is bound to be really 'fun'.
<2> I'm dealing with a 90GB database with 183 tables AND NO INDEXES
<3> well at least you knew thbat was going to be a whorry bitch
<3> this was supposed to take 5 ****ing minutes
<6> ****s
<2> my boss expects me to come up with this **** in 5 minutes
<6> cubert, how did you inherit 90GB of data with no indices?
<2> I'm just reporting from the data
<2> it's our ERP system
<6> the ERP system has no indices?
<2> the ****tards who created it did it originally way back on System 36's, then "upgraded" through the past 30 years
<2> here's an example of the SQL I have to deal with, it'll probably look familiar: http://sql-servers.com/nopaste/?show=322
<3> ---------------------------
<3> SQL Server Enterprise Manager
<3> ---------------------------
<3> SQL Server registration failed because of the connection failure displayed below. Do you wish to Modify anyway?
<3> To connect to this server you must use SQL Server Management Studio or SQL Server Management Objects (SMO)
<3> ---------------------------
<3> Yes No
<3> ---------------------------
<2> I think you've helped me with this
<4> is there an easy way to say "select field7 of max(field2)" or something?
<4> so you use field2 for the max, but it returns field7
<3> ohhh the bastards
<3> like select max(isnull(field2,filed7)) ?
<4> i know it can be done via joining against itself
<4> no.. like return field7 on record containing max(field2)
<4> i don't care about max(field2) except for the max comparison... i WANT field7
<3> isnt that going to give you some aggregat error ?
<4> yes!
<4> which is why i'm asking
<2> heh
<3> lol
<4> i can do select x where PK in (select PK,max(field2)) or something
<4> but was hoping to avoid the subselects since it *should* be a fairly simple operation
<7> :O
<7> i
<7> i was kicked!
<2> why do you care that max(field2) is returned? can't you just return it and ignore it?
<3> dang
<3> you're on a roll - kicked and killed all in the same day
<4> cubert: hm
<2> you're only going to be getting one record anyway, right?
<4> yea
<2> or do you want all of Table1 that matches max(field2) in Table2?
<4> no just the one record
<4> was thinking cuz it was joining... but i can prolly just put JOIN target ON MAX(field) or somethin
<3> well sql 2k5 can't push lol
<2> SBrick98: or use a CASE
Return to
#sql
or
Go to some related
logs:
#unixhelp
yum segmentation fault parsing package install arguments
#flash
#windows
#politics
#beginner
#beginner
Toress from argentian
#nhl
suse check-link-speed
|
|