#mysql
01 November 2007
Total 22 pages. You are browsing page 3/22.
First :: Prev :: [1] [2] [3] [4] [5] [...] :: Next :: Last
01:51 <****> I think it's a php programming question, personally.
01:52 <****> I do agree it's really hard to decipher.
01:55 <****> is mysql always a transactional database or does it depend on if you use InnoDB, etc.
01:56 <****> profounded: myisam does not support transactions
01:56 <****> yes, it depends on the storage engine
01:56 <****> innoDB does.
01:56 <****> ok, so lets say im using InnoDB with php, do I have to commit every time?
01:57 <****> you don't have to create transactions.
01:57 <****> depends on your autocommit setting
01:59 <****> thank you
02:05 <****> i have mysql running right now but i don't know where the socket is, how can i find where the socket is located?
02:06 <****> JakeConnor: knowing where the socket is is likely not your real issue.
02:06 <****> JakeConnor: what issue are you trying to resolve, exactly?
02:06 <****> i just need to know where my socket is, thanks
02:09 <****> JakeCOnnor: netstat -napt
02:09 <****> nanus: netstat: t: unknown or uninstrumented protocol
02:09 <****> JakeConnor: Check the error log of mysqld, it will tell you
02:10 <****>
02:11 <****> _k: you are right, it did tell me! thanks :)
02:12 <****> JakeConnor: or run 'ps aux | grep mysqld' and see if it tells you.
02:13 <****> its ok i got it now thanks
02:15 <****> hi there
02:15 <****> anybody have tried mysql over glusterfs?
02:15 <****> iamamoron: what's glusterfs?
02:15 <****> www.gluster.org
02:16 <****> it is for clustering
02:16 <****> problem with mysql right now is clustering
02:16 <****> iamamoron: interesting... i am trying out some clustering ideas myself
02:16 <****> it is memory based
02:16 <****> what do you mean JakeConnor?
02:16 <****> iamamoron: not completely true... with 5.1 it doesn't all have to be in memory
02:17 <****> yes in 5.1 but indexes cannot be saved in hard drive
02:17 <****> it is still memory based
02:17 <****> iamamoron: what do i mean about what?
02:17 <****> also 5.1 is not for production yet
02:17 <****> iamamoron: yeah indexes are the only thing that must be stored in memory but you can store the rest of your data on disk
02:18 <****> what i am trying to do is clustering
02:18 <****> i think the whole index column has to be in memory
02:18 <****> iamamoron: 5.1 will be released any day now... i'm already using it in production
02:18 <****> nop the indexes in none clustered mysql resides in hard drive
02:18 <****> ?
02:19 <****> iamamoron: well the word clustering is very vague... if you are trying to do mysql clustering then you have to store indexes in ram, no other way around it
02:19 <****> also 5.1 is not for production yet
02:19 <****> jbalint
02:19 <****> jbalint: pm?
02:20 <****> its worth giving 5.1 a shot anyways since its going to be production any day now
02:20 <****> deadroot: hey
02:20 <****> i dont think so
02:21 <****> ok, i guess this conversation ends there then heh
02:21 <****> what was the variable to set the recursion limit named again?
02:21 <****> can't remember, can't find it via google
02:23 <****> /home/back.sql.gz" How would I log the output if any, so I know if the command fails for some reason.
02:24 <****> mikefoo: try #bash?
02:24 <****> I did, heh
02:24 <****> only capture stdout. stderr will still print to the console
02:25 <****> ToeBee: ok, so how do I capture that to a file?
02:26 <****> mikefoo: google redirect stderr
02:27 <****>
02:27 <****> I tried that, no work..
02:32 <****> Question, im getting over 341 connections a second on my db, and the application is losing connection to the db through the unix socket. I've enabled query caching, but im still seeing a huge amount of DMS on select, any ideas on how to stop the database errors?
02:36 <****> anybody have tried implementing mysql over glusterfs?
02:39 <****> http://rafb.net/p/F82ztL43.html I'm getting an error on line 8. But I can't see what it is.
02:42 <****> Facedown2: is mod a reserved word?
02:42 <****> Maybe..
02:42 <****> it is
02:42 <****> put backticks around it `mod`
02:43 <****> Facedown2: but I don't recommend using reserved words in any event
02:43 <****> http://dev.mysql.com/doc/refman/5.0/en/reserved-words.html
02:43 <****> I see
02:43 <****> I'll change it to moderators then, thanks.
02:50 <****> will mysql ever implement lucene as their full-text replacement anytime soon?
02:56 <****> how do you get rid of binlogs? is it safe to just wipe those files?
02:56 <****> i already disabled binlog in conf but they are still there taking 5 gigs of my hd space
02:58 <****> robboplus_: search the manual for 'purge master files'
02:58 <****> ki77a77 i haven been up since 16 hours.. can't you just tell me please?
02:58 <****> i can hardly touch the mouse already
02:59 <****> have*
02:59 <****> robboplus_: then read it tomorrow. http://dev.mysql.com/doc/refman/5.0/en/purge-master-logs.html
03:00 <****> can oracle perform better searches than Lucene?
03:01 <****> unbeatabl: there is a similar indexing called intermedia
03:03 <****> thanks ki77a77
03:03 <****> np
03:04 <****> can i somehow sum certain int column from all rows
03:05 <****> select sum(certain) from table;
03:09 <****> ki77a77 i did PURGE BINARY LOGS TO 'mysql-bin.000095';
03:09 <****> and it didn't remove ANY file
03:09 <****> while there are tons of them and 95 is the last one
03:09 <****> what am i doing wrong?
03:11 <****> see why i hate manuals
03:11 <****> it fucking doesn't work as expected
03:11 <****> and where are you now?
03:11 <****> all you can help with is tell me to "rtfm"
03:12 <****> old good rm works best
03:16 <****> I'm thinking about setting up a two machine cluster, from what I have seen it should be possible to install data, sql and management nodes on each box although only one management node would be active at any given time, has anyone here tested this kind of setup or is there something that would advise against such a setup ?
03:34 <****> robboplus_: how long did the command last for?
03:35 <****> ie, take to run
03:35 <****> Dallu1: just 2 machines is not suficient.
03:35 <****> Dallu1: The management server needs to be on a seperate machine, for fully redundance
03:36 <****> and, er, isn't "purge master logs" what you want?
03:36 <****> ahh, same deal
Total 22 pages. You are browsing page 3/22.
First :: Prev :: [1] [2] [3] [4] [5] [...] :: Next :: Last