| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Comments:
<0> would it merge the indexes?, and if you're wanting to order by that date, then it might take longer, no?
<1> Jivedue: Here it is in a nutshell: If you need it you will love it, if you don't then ignore it.
<2> !tell us about 5.1
<3> shabbs asked me to tell you this: 5.1 is planned to support table partitioning, disk-based Cluster, row-based replication and more. More details as the release moves towards beta.
<4> keylogger hello!
<0> well, I can understand how it could be useful
<1> A true VLDB will benefit. If you have to ask if it will take longer then you may not need it.
<0> just I like to really pick at things
<4> startkeylogger hello!
<2> Therion: tag
<1> bugjodepingha: What in the world is that?
<1> Aah, thank you Therion.
<0> haha
<5> :)
<6> shabbs: Anything else?
<2> mhillyer: norton bug exploit
<2> Therion: no, that's it. thank you :)
<0> mhillyer: yah, I'm reading that article, the m***-delete benefits sound great too
<7> alphablue: ok you shouldn't need to down apache to run the script.
<1> Well, time for lunch.
<7> alphablue: and if you're using mysql 4.0 or higher you don't need to restart mysql for anything less than startup "options" (not variables) and files.
<8> todo: it's a forum, so i wanted everything to be 0
<8> and i didn't wanna turn off mysql and not apache, people would be more pissed
<7> "flush status"
<8> i know, i still like to make sure things are there for next time
<7> ok
<8> Current tmp_table_size = 128 M
<8> 95% of tmp tables created were disk based
<8> Perhaps you should increase your tmp_table_size
<8> i knew this was gonna happen, :)
<9> Would it be possible to specify a constraint (like ENUM) that would allow any value but 'X' and 'Y'?
<8> it has to do with invision board having ****ty queries
<7> alphablue: that's why I say "perhaps"
<7> "Perhaps" you can fix it another way too
<8> i know, not saying you're wrong.. but just people may see recommendations and notice things don't change ..
<7> yea, it's intended as a 1st step.
<7> in depth tuning is much more of an art than a script can do.
<8> i'll leave it alone for 48 hrs see what happens, we get around 5 million pageviews/day should fill up the cache etc very well
<7> cool
<10> TodoInTX: where is this script?
<7> in your case if it's a normal day with plenty of queries then you can probably get away with 24hrs.
<8> todo: indeed, we've been working with the ipb people, 2.1 is gonna have much better performance.
<7> http://www.day32.com/tuning-primer.sh
<10> thank you :)
<8> if the box dies due to out of memory issues, i may lower some of the settings :)
<7> certianly.
<7> in some cases it will look of over allocation (like query cache size)
<8> yep right now it's saying 32MB is too high
<8> but prior to restart it said 16MB was too low
<7> that's why you have to give it time.
<10> TodoInTX: very nice.
<8> hmm when i use the misc option, it says i have 72 slow queries, it'd be nice if you did similar to query cache, where you have slow query ratio, so you can see 72 slow queries out of 70,000
<7> infi: I may rewrite the thing all over again in perl so that it will accept user= p***word= host= stuff from getopts.
<10> TodoInTX: there is a getopt(1) ;)
<10> I don't wish it upon my worst enemy, but it exists
<10> perl is fairly standard though, and crash-me runs in perl, so it seems reasonable
<7> alphablue: I find that the ratio is less important. I try to shoot for 0 slow queries.
<8> me too, but it's more of a snapshot for the average joe
<7> if someone see oh... I have .1% slow queries they think I'm ok!
<7> it's not true/
<8> ok, then change to you have 72 slow queries out of 70,000 it's more of a way for me to judge i guess. sometimes like in my situation, i know the next release will drop in slow queries tremendously, but i have to wait until it p***es QA
<11> I want to select from multiple tables, but sometimes one of the tables doesn't have data pertaining to the primary key. When that happens, no result is returned for that entry. How do I get around that?
<12> I thought MyISAM tables didn't support foreign keys? Why did msyqld just access some SQL that created a foreign key? And, the foreign key is in the database. I can view it in MySQL Query Browser.
<7> alphablue: ok done, get latest version.
<10> TodoInTX: actually, alphablue has a point. for all of the items that you recommend changes for, you should show the current setting (and ratio, if appropriate/available)
<10> imho
<8> todointx: nice
<12> Do MyISAM tables accept foreign keys in MySQL 4.1? The docs say they don't.
<8> 115 out of 100,00k queries
<5> if the docs say they don't, then trust the docs
<7> don't I?
<7> infi: did I miss one ?
<8> you do except in the slow query thing, which you fixed
<10> yes, now they do :>
<12> Jeewhizz: but, I've got them in the database and can view them with QB. I just haven't checked if they work yet.
<10> phew, 808 lines of bash scripting.
<7> hehe
<7> it's my hammer
<10> karate works wonders ;)
<10> nice job though, very worthwhile.
<7> the perl should be much shorter. I won't have to pop back and forth to the database everytime I want a value.
<10> that's what I was thinking too
<10> you could even combine them, write a test for `which perl`, and then generate the perl on the fly with a cat << EOF ;)
<10> but that would just start getting nonsensical. heh
<7> uh.. yea
<13> is there any other tool for comparing databases like red gate`s sql bundle ?
<7> sqlyog has something like that.
<13> hm i need tool that can compare mysql but also ms sql
<7> RockGalama: not sure, maybe the migration tool ?
<7> http://dev.mysql.com/downloads/migration-toolkit/1.0.html
<8> http://radar.oreilly.com/archives/2006/03/entrepreneurial_proverbs.html
<14> guys, I want to create a WHERE clause based on a date ....
<8> nice read for any other entrepeneurs in the house
<14> i.e. the date is held 2006-01-01, I want to query my db, for example, where the "day" is sunday
<14> can i achieve this?
<7> infi: those tune your db in 5 min or 3 min post on planetmysql pissed me off.
<13> thx TodoInTX
<13> what is difference between mysql and maxdb ?
<7> RockGalama: lots.
<14> can anyone help?
<7> !m rusa date and time functions
<3> rusa: (Date and Time Functions) : http://dev.mysql.com/doc/mysql/en/Date_and_time_functions.html
<13> maxdb is better faster .. where can i find differences to read about it
<14> i just don't know how to construct the clause
<7> rusa: hold on I'll make you an example.
<14> i.e. ...
<14> WHERE DAYNAME('1998-02-05') = 'thursday'
<7> RockGalama: http://dev.mysql.com/doc/maxdb/pdf/whitepaper.pdf
<14> ?
<13> thanks
<14> or, WHERE DAYNAME(t1.date) = 'thursday'
<14> is that correct?
<7> yes, that will work or this can -- select * from table where date_format(timestamp, '%W') = 'Sunday';
<7> I think you need to capilatize Thursday though.
<14> oh..
<14> thks
<14> :)
<9> Hm. I mean. -would- it be possible to specify a constraint (like ENUM) that would allow any value but 'X' and 'Y'? Or is that just impossible and should I keep that logic in my SP? (:
<7> jasper: I've not heard of a method for doing this in SQL.
<7> jasper: not to say that one may not exist
<15> hi there! Im using php, how can I create a table row with data instead of getting a mysql error when querying the data thaty does not exist?
<2> jasper: considering that it's already in your SP, it's already considered 'database level' so I wouldn't worry about it
<0> jasper: in your CREATE TABLE you can CHECK the value of columns
<7> carlos-the-man: so you're trying to do an update but failing that you want in insert ?
<15> I have dafault data in case it does not exist :)
<15> is this possible?
<7> I still don't understand what you're asking for exactly.
<16> I want to replicate a single database across to a backup server...there's other tables that exist on the machine I'm backing up to, but I would like it to be a "slave" for only 1 database...
<16> can I do it?
<7> OneFix_Work: yes use the "replicate-do-db=" option
<7> on the slave.
<7> if you don't use the "binlog-do-db=" on the master the slave will get all updates then filter using the replicate-do-db option.
<15> UPDATE groups_to_products set access=1 where products_id=30 and customers_group_id=4
<15> TodoInTX this is the db I use to relate products to customer groups, and am working on the part where I set permissions and the data is not on the table, so Id like to populate the DB automatically in case it does not exist
<15> "UPDATE groups_to_products set access=1 where products_id=32 and customers_group_id=8" does not exist, so Id like to have this data created instead of the ugly error :D
<16> ok, so what about doing it for say 2 databases on the master
<16> Can I give replicate-do-db a comma seperated list?
<7> OneFix_Work: read all about it here: http://dev.mysql.com/doc/refman/4.1/en/server-options.html
<7> sorry here: http://dev.mysql.com/doc/refman/4.1/en/replication-options.html
<7> carlos-the-man: can you get away with insert first , if duplicate update ?
<7> http://dev.mysql.com/doc/refman/5.0/en/insert-on-duplicate.html
<7> hello?
<2> echo echo echo
<8> ARGH i hate truechange
<17> anyone know if altering a TEXT to a BLOB will alter data stored in those fields
<18> nope
<19> hi
<17> I have a problem suddenly with people pasting text from MS Word On a mac into TEXT fields
<7> FreemaniaX: the only differrence between those that I can see is that one is measured in bytes (BLOB) and the other in characters (TEXT);
<7> if you have unicode characters filling to almost the capacity of TEXT then the data could be truncated when you convert to BLOB.
<18> ****ty MS Word data can be solved by converting character sets
Return to
#mysql
or
Go to some related
logs:
libmatroska für suse10.1
mount error 1 = Operation not permitted mount.cifs
#linuxhelp
roaming Id server openldap
#perl
openembedded glibc amd64
kxdocker gnome ubuntu
perl MSocket SOMAXCONN
xubuntu mplayerplug-in
linux turboprint torrent
|
|