| |
| |
| |
|
Page: 1 2 3 4
Comments:
<0> yes, use Enterprise Manager
<0> if you mean "easy" as in "let me queue up all the restores and walk away while it is doing it", then yes, use SQL
<0> RESTORE DATABASE is the command to look up
<1> Ok thanks
<1> I was looking for queue up all the restores
<1> LVK: Working great... Thank you for your help!
<2> how do you clear a transaction log in SQL Server 2k5?
<3> Odyss3us, back it up and that allows it to free up the old transactions
<3> if you can set it to simple that makes it use a circular buffer. then set the size to a reasonable limit.
<3> then issue DBCC SHRINKFILE(DatabaseName_log, 2) where 2 is the size you are targeting
<4> circular buffer?
<2> cool thanks
<3> i till replace old transactions with new ones if under options you set the recovery model to simple
<3> that lets no control the growth of the log file and make sure it never gets full
<3> but back it up and shrink it first
<4> i have to say i don't agree with this advice
<3> depends on the need.. if you have a transactional type where rollbacks are common or important, so not use simple
<4> at least without giving caution to the consequences of not being able to do tran log backups
<3> sorry, you are correct
<3> it was kind of tip of the iceburg advice
<2> its a dev database, I really don't care about the contents of the log
<3> in properties under files set the size of it once you shrink it
<5> hi can we join two tables from different dbs?
<3> testdrive: they were talking about that here the other day and the answer was yes, sorry I don't remember the syntax.
<6> testdrive, what dbms?
<7> testdrive : which rdbms? are the dbs on the same server? running under the same rdbms instance?
<4> whoah
<6> sweet
<6> I hate when people do that
<6> Who's going to the SQL Conference?
<4> not me this year
<7> nor me - wrong continent
<6> It's so odd; the welcome msg hypes the conference, but apparently nobody on channel is going
<8> how much does it cost?
<8> i wouldnt mind going and i can get the company to pay for it
<8> and orlando is about 2-3 hours from me
<8> I'm only interested in DR and Performance breakouts
<2> so I have a proc that took about 5-10seconds to run on SQL Server 2k, I backed up the database, uninstalled SQL Server 2k and installed 2k5, restored the database and set it to SQL 2k5, the same proc has been running for 7min with no results
<2> what possibly could have gotten ****ed?
<8> indexes?
<8> is this the first time you've run it too?
<8> nothing is cached if so
<2> on the new server, yes
<8> look at the execution plan, it'll tell you whats taking a while
<2> doesn't it have to execute before I can do that?
<9> it shouldnt
<6> Rizard, it's about $1400
<6> but it starts tomorrow
<4> Odyss3us check data types
<4> sql server 2000 sp4 and sql server 2005 handle stuff differently
<6> I believe the SQL 2000 statistics are invalidated upon upgrade; you must run UPDATE STATISTICS WITH FULLSCAN
<6> on all talbes
<4> especially if you're comparing two different data types (like numeric (35,2) to a numeric (19,1), etc)
<4> plus if you have a "Not in" that will get handled differently as well
<2> is there a proc that updates stats on all tables
<6> sp_updatestats
<2> cool, thanks
<2> well thats certainly doing something :)
<4> after sp_updatestats you need to dbcc freeproccache or sp_recompile <table> to get a new plan
<2> on what tables?
<6> I still say you are better off running UPDATE STATISTICS [tableName] WITH FULLSCAN
<4> just one table used in your sp
<6> that way you get a new set
<2> any table?
<4> any table used in your SP
<2> cool
<2> hmmm proc is still taking a while
<2> at 1min so far
<6> the sp_recompile is unnecessary; any time a table has its statistics updated, all plans referencing that table will be flushed from cache
<6> Odyss3us, how big is this database?
<2> ~500MB
<2> this is a very hard core proc
<2> anything else you can think of?
<2> or might it just take a looong time its first run
<6> it will take however long it takes, considering that you didn't use UPDATE STATISTICS on the relevant tables; it is running on all user-defined tables
<2> ah
<4> Odyss3us did you read what i said before?
<6> shouldn't take very long on < 500MB data though
<4> you seem to ignore every other thing someone says
<6> ;)
<2> I did what you said Mixx
<4> not that
<4> are your data types the same?
<2> yes
<4> are you using NOT IN?
<2> possibly
<4> paste your code and table structure to topic
<2> you really don't want to dig through that
<4> i didnt say I would
<4> but someone might
<4> ;)
<2> http://pastebin.com/635142
<2> don't say I didn't warn you
<4> where's it getting stuck?
<10> Odyss3us : use the right paste site, dummy
<10> Odyss3us : read the channel topic again
<2> hrh
<2> er heh
<2> I'm really not sure
<4> ................
<4> run execution plan
<6> First off, is the machine done regathering the stats?
<2> yes, took about 5 min
<2> hmmm it seems to be semi-frozen
<4> yeah
<2> as in SQL Management Studio isn't responding
<4> it does that
<4> yes
<4> when you turn execution plan on it does that
<6> I swear; you don't listen very well. YuppieScm warned you to use the right paste site
<4> he doesn't
<2> will it take care of itself before this query ends
<4> no
<2> I haven't pasted anything since he did
<2> or it, or whatever you like to call your bot
<2> great, since this query hasn't successfully executed yet
<4> the point was, to paste what you pasted to the wrong site to the right site
<4> thank you
<4> jesus
<2> http://sql-servers.com/nopaste/?show=498, happy?
<4> uh oh
<2> I told you didn't want to see this
<6> you're right
<4> Odyss3us it' snot like someone's going to pour through the whole thing
<4> but when you tell us which part is taking up the most resources we'll be able to figure it out
<7> that's a ****ing ugly SP
<2> yeah I know
<4> which is getting exceedingly difficult to do
<2> I think it's the INSERT INTO @actTab
<4> since yer kind of a pain in the ***
<4> ;)
<7> did no-one everr tell you that returning more than one result set from an SP is a sign of poor design?
<2> I'm really sorry, I'm a bit cranky
<2> its faster than calling several SP's in different connections?
<2> \s\?\
<2> espically when the result sets are inter dependent
<7> [02:45:17] <2> its faster than calling several SP's in different connections? <= is that a question or a statement? have you benchmarked it?
<2> I did \s\?\, I dumped the question mark
<2> its definatley faster, and we did test it
<2> in 2000 anyhow
<2> I don't understand why this isn't working... I mean what changed?
<2> I mean the process I did to upgrade it was correct, no?
<4> did the sp finish?
<2> no
<2> :(
<4> are you letting it?
<2> yes
<4> well when it finishes then we'll know
<2> somethings really ****ed *sigh*
Return to
#sql
or
Go to some related
logs:
unable to join channel (address is banned)
maxscript imgtag
Ubuntu, memtest, how long
hintia
remove + movicliploader
#politics
mIRC buggie
lundqvist
#firebird
#windows
|
|