Re: Yet a solution for the No-Indexes-Used-In-Selects-With-Or-In-Where-Clause-Problem

Michael Braun wrote: > > I'm using mysql 3.23.41 on MacOSX... > The problem is, that MySQL isn't able to use indexes for OR in the WHERE > clause yet and therefore the additional OR prevents optimal use of > indexes. MySQL uses only one index for a select, so it can't use an index > when there is

Re: Slow results with SELECT's that return large numbers of results

Jeremy Zawodny wrote: > > > > SELECT COUNT(*) FROM foo WHERE ... > > > > > How can I get this to work with a GROUP BY? > > Can you post the query along with the output of EXPLAIN so we > can get a good idea what's going on? Query: SELECT documents.document_id FROM documents, dictiona

Re: Efficient Boolean combination of results of several SELECTs

Sinisa Milivojevic wrote: > > No, but in 4.1 there will be INTERCEPT and MINUS too, beside > UNION's (with or without ALL option). Those three combined can > produce any of the above relational operations. wooah! *dribble*! :-) When can I start using this code? Where can I get it from? ch

Re: Slow results with SELECT's that return large numbers of results

Jeremy Zawodny wrote: > > SQL_BUFFER_RESULT (http://www.mysql.com/doc/S/E/SELECT.html) my help, > depending on what the real bottleneck is. H... I thought temporary tables were bad? > Sure, change your > > SELECT * FROM foo WHERE ... > > to > > SELECT COUNT(*) FROM foo WHERE ... > >

Re: Efficient Boolean combination of results of several SELECTs

Jeremy Zawodny wrote: > > > Yup, but given that each WHERE clause is already pretty hairy, I was > > hoping someone would say "oh, sure, here's some set operations to > > complement UNION" ;-) > > How about, "try out MySQL 4.0 with UNION support"? :-) Can UNION do boolean AND, NOT and OR combin

Re: Efficient Boolean combination of results of several SELECTs

Jeremy Zawodny wrote: > > Hm. It sounds like you just want to combine the WHERE clauses > (intelligently) and end up with one big query, right? Yup, but given that each WHERE clause is already pretty hairy, I was hoping someone would say "oh, sure, here's some set operations to complement UNION

Re: Slow results with SELECT's that return large numbers of results

neeraj arora wrote: > > Hi, > I think indexing the column(s) can help you. Already spent a week figuring how to get the SELECT to work fast when a few results are returned, os all appropriate columns are indexed ;-) > u can find more on this page > http://www.linux-mag.com/cgi-bin/printer.pl?i

Efficient Boolean combination of results of several SELECTs

...in MySQL, to keep the list filter happy :-( Hi, I have several different SELECTs, each of which returns a set of results which I'd like to combine in a boolean fashion. So, say I have 4 selects, w, x, y, and z, I might want to do: ((x AND y) OR Z ) AND (NOT w) What's the best way of going

[ot] harsh mail rejection

[EMAIL PROTECTED] wrote: > > Your message cannot be posted because it appears to be either spam or > simply off topic to our filter. To bypass the filter you must include > one of the following words in your message: > > database,sql,query,table Why does this list not include SELECT and INSERT?

Slow results with SELECT's that return large numbers of results

Hi, If I have a SELECT that returns a large number of results, say > 1000, then returns the results for the SELECT takes much over an order of magnitude longer than if the results set had been, say, 100. Now, is there any way I can quickly return the length of the result set withotu actually ret

MySQL still hanging on Win2K :-(

Hi again, Sinisa... Chris Withers wrote: > > Okay, table_cache increased from 64 to 256... > > > * check free disk space on all drives Well, it got up to indexing 10,000 documents this time but then hung again. Now, I don't know if this is because increasing the table_cac

Re: MySQL thread hanging on Win2K

Hi Sinisa, Sinisa Milivojevic wrote: > > There are several possibilities. You could be reaching some OS limits > or haning user / group permission problems. Don't think its permissions, since I had a similar process which ran fine, but I had to change the schema to use the last insert id for ot

MySQL thread hanging on Win2K

Hi, I've got a problem with 3.23.44-max-nt on Windows 2000 using Berkley tables. I'm running an import process which inserts a big set of rows ('bout 27 million), as part of a document indexing process. Each new row uses the AUTO_INCREMENT_ID from the previous row. So, I have a loop which basica

Re: InnoDB select slow too

Heikki Tuuri wrote: > > Chris, > > what is your my.cnf or my.ini like? Do you have a big enough > innodb_buffer_pool_size? set-variable = innodb_buffer_pool_size=300M big enough? Chris database, sql, query, table (just to keep the stoopid list software from whinging..) -

Re: InnoDB select slow too

[EMAIL PROTECTED] wrote: > > Your message cannot be posted because it appears to be either spam or > simply off topic to our filter. To bypass the filter you must include > one of the following words in your message: > > database,sql,query,table > > If you just reply to this message, and incl

Re: InnoDB inserts: HELP! :-)

Joe Ellis wrote: > > Just For My Info: does it take long for one insert. if so, what does > the insert statement look like. the reason i ask is becuase i use > InnoDB and inserted a few records and it was pretty quick. but maybe > thats becuase my statement was: > insert into table (user,email

Re: InnoDB inserts: HELP! :-)

Norman Khine wrote: > > Welcome to the MySQL list;^) good to see you here... *grinz* It's weird being a newbie again :-P > "In inserts InnoDB uses the insert buffer to merge secondary index records > to indexes in batches. That saves a lot of disk i/o. In rollback no such > mechanism is used, a

InnoDB inserts: HELP! :-)

Okay, second time of asking, does anyone know why InnoDB inserts are an order of magnitude slower than BDB inserts? Is this true only for me? If so, what am I doing wrong? Is it a FAQ? If so, where can I find the answer? Are InnoDB tables supposed to behave like this? If so, why? I know I'm n

InnoDB inserts slow?

Hi there, Was using BDB tables for some full text indexing schema I've got. Indexing a document involves inserting a whole lot of rows in one table, having done some selects and possibly inserts on another table. With BDB, this was averaging 1.5 seconds a document. Thought I'd try InnoDB since