#mysql
15 November 2007
Total 10 pages. You are browsing page 3/10.
First :: Prev :: [1] [2] [3] [4] [5] [...] :: Next :: Last
02:18 <****>
02:18 <****> i also tried synaptics
02:19 <****> try sudo apt-get reinstall --purge mysql-server-5.0
02:20 <****> that command gives me an error about reinstall not been a command
02:20 <****> sorry sudo apt-get --reinstall purge mysql-server-5.0
02:21 <****> got it i did a sudo apt-get remove --purge mysql-server-5.0
02:22 <****> the install it and it worked thanks
02:22 <****> ok that's fine too
02:23 <****> ok thanks now i have to set it up, have to do some reading thx again
02:23 <****> you are welcome
02:26 <****> hi - I just encountered the bug shown here: http://www.delphifaq.com/faq/databases/mysql/f1110.shtml
02:26 <****> can anyone give me a status on that bug? I had to change some code when MySQL got updated
02:28 <****> when using select ... into outfile, how can I include the name of fields to be exported as well?
02:35 <****> question on licensing...if i have open source but commercial website using mysql, its free right? under GPL
02:35 <****> yup
02:36 <****> and if its closed source, I must pay?
02:36 <****> mysql community edition(the one with distributions and you can download don't come with licensing. You have to contact mysql for the paid for version
02:36 <****> ok
02:37 <****> and while there are some few differences, it is the same thing. you are paying for support from mysql and helping a great company
02:47 <****> Hi! I use a Xeon SMP box (2 dual core CPUs) 3ware 9550S, with Debian AMD64 and MySQL 5 (/testing). The performances are awful: loooong execution time for queries while nearly no CPU nor disk activity. thread_concurrency=8 and innodb_thread_concurrency=4 (to cope with known MySQL-on-such-SMP-system-related problems). Is somebody using a similar setup without sorrow?
02:48 <****> natmaka , yes
02:48 <****> Yay :-)
02:48 <****> Is there any trick?
02:48 <****> but not completely
02:48 <****> Even a partial solution will do, I'm desesperate
02:50 <****> hi, I have a data base for news, when each atricle there is creation_date which is a TIMESTAMP and the value is inserted then the article is created as NOW(), my question is how can I show only the articles which is less than 18 months in my query?
02:51 <****> NOW()
02:51 <****> natmaka, feed mysql lots of memory :)
02:51 <****> Dauby Done. 9 GB
02:51 <****> but you have to do it with tuning. just putting memory in won't make it effect
02:51 <****> for 12 GB total on the box
02:52 <****> Tuned with tuning-primer.sh
02:52 <****> Alloos1 you may use INTERVAL
02:52 <****> is that a reporting script?
02:52 <****> natmaka: how would the query look like ?
02:52 <****> a tuning-helper
02:53 <****> those are misleading sometimes depending on how often mysql is reloaded
02:54 <****> natmaka: I think I found it in manual, will go check it out
02:54 <****> Alloos1 : you will prolly have some 'where' condition in the "IF(creation_date<DATE_FORMAT((DATE_SUB(NOW(),INTERVAL 180 DAY)...
02:54 <****> form
02:54 <****> kinda odd low disk activity
02:55 <****> how much IO wait?
02:55 <****> Dauby http://www.day32.com/MySQL/tuning-primer.sh
02:56 <****> Dauby never above 30k reported by vmstat, albeit the very same vmstat shows up to 280k under bonnie++ and 200k with a 'dd' running
02:56 <****> Dauby nearly always below 10k
02:57 <****> how many queries per second, natmaka
02:57 <****> no more than 700
02:58 <****> mainly selects
02:59 <****> may be tmp_table_size and heap_table_size
02:59 <****> reported thx to munin, during the last year of activity. No awful locks (since I limited the max simultaneously active threads)
03:00 <****> nope, less than 10% of queries
03:00 <****> I get around 1k per sec eating 350% cpu and almost 0 io
03:00 <****> key buffer?
03:01 <****> Dauby: wow, not bad!! mainly selects?
03:01 <****> larger than total keys?
03:01 <****> yeah selects galore
03:01 <****> 128M
03:02 <****> wanna try a different tuning script?
03:02 <****> yep, thx
03:03 <****> http://rackerhacker.com/mysqltuner/
03:03 <****> from a dba at rackspace
03:03 <****> thx, I will try it!
03:03 <****> right now
03:03 <****> your key buffer seems kinda low
03:03 <****> mine is set at 584M
03:04 <****> it is a perl script
03:04 <****> so you gotta perl mysqltuner.pl
03:04 <****> Hello guys, quick question... I have a table with the following rows [ id(int), sid(int), string(varchar512) and lang(tinyint)]. I need to find rows that have lang=1 but missing lang=2 based on sid value. Any ideas if I can do that in SQL?
03:05 <****> Dauby I mainly use InnoDB, with a single big ISAM table
03:05 <****> you are keeping your keys up to date with mysqlcheck -a db ?
03:06 <****> It just seems to be that getting all the SID, lang=1 then checking against the table for lang=2 would be a huge ressource hog because I have over 6000 rows and I don't want to make 6,000 queries!
03:07 <****> ionix, one sec
03:07 <****> Dauby Nope, I thought... you mean, I have to do it even if I only inject contents then do a few updates ?
03:08 <****> mysqltuner.pl seems cool...
03:08 <****> thx!
03:08 <****> :)
03:09 <****> ionix, you want it all to return in a single query?
03:10 <****> hmm I can loop it
03:10 <****> I was just worried about the performance issues
03:11 <****> Basically, this table contains strings in both English (Lang=1) and Japanese (Lang=2).
03:11 <****> natmaka, mysqlcheck will make mysql choose the right keys and make it make the right decisions
03:11 <****> I am developping the app in English and someone else will translate my strings from English to Japanese so I am building them an interface to show strings that need an English translation
03:12 <****> Dauby OK, I snooped around using statistics and invoked mysqlcheck a couple of times, but will run it right now
03:12 <****> I have an other table called "stringid" which contains id,name. And the id in that table is the sid in the string table that I've described above
03:12 <****> ionix, I could be wrong but it seems like you could do two simple selects and get what you want
03:13 <****> How can I retrieve the row that I just inserted?
03:13 <****> Dauby do you periodicaly (cron?) run mysqlcheck?
03:14 <****> from table where Lang=2;
03:15 <****> natmaka, I have been meaning to but I usually do it by hand just to eyeball the results. doesn't take long;
03:15 <****> I'm inserting a row with a Unique Index by doing SET uid = NOW() + 1, and in my code I'd like to know what that uid is. Is this possible?
03:15 <****>; (??)
03:15 <****> Dauby thanks, will do
03:16 <****> Anyone?
03:17 <****> ldlework repeat.
03:17 <****> yes, its possible, but I don't remember how. php? there is like a mysql_last_inserted(not exactly this function name)
03:18 <****> last_insert_id?
03:18 <****> www.php.net/mysql_insert_id
03:18 <****> When I create a new row under my 'job' table, I set the Unique Index by SET = NOW() + 1 to set it to the timestamp. I want to know the timestamp that MySQL generated so I can refer back to this row.
03:18 <****> i know of a function called mysql_insert_id
03:18 <****> mysql_insert_id returns the last insert which resulted in an Auto Inrcement.
Total 10 pages. You are browsing page 3/10.
First :: Prev :: [1] [2] [3] [4] [5] [...] :: Next :: Last