| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Comments:
<0> (if travelling across the internet)
<1> How to make it listen on ext IP adr?
<1> What about unsafe?
<0> the mysql protocol is not really encrypted, and having the daemon listening to the internet at large invites "unwelcome guests" trying to knock on your door.
<1> OK.
<0> at least filter hosts to only allow the exact hosts you want to query it from.
<1> ok, thanks :)
<2> Hi all
<2> I have a database with names. Now I want to create for every record a unique number, and I want this number randomly to be choosen. Ofcourse this can be done using a small piece of code, but perhaps there is a more easy/efficient way to accomplish this with mysql?
<3> can some one tell me is there any way to replicate some tables from mysql database instead of replicating whole db ?
<3> any one ?
<4> <12> I am creating a lonf string in which I do searching and replacing of characters
<5> how does one use count in sql
<0> select count([column|*]) from [table] where [...]
<6> Does version 4.1 support locking on individual rows or do I have to lock an entire table?
<7> ctooley: it depends on the table type
<7> and not the mysql version
<7> innodb does row level locking
<7> myisam table level locking
<7> berkley db page level locking
<6> so if we were to switch to innodb we could do row level locks.
<6> thanks
<8> when i changed to innodb i started getting "deadlock found" for a certain delete statement virtually every time
<8> delete from table1 where not exists (select something from othertable where table1.x=othertable.y)
<7> i dont think subqueries are supported in delete statements
<7> if the table from which you delete is included in the subquery
<6> I want to do something like "update table set handled = 1 where id = ( select id from table where handled != 1 order by id desc limit 1)" and I would like the id of the column it just performed that update on.
<6> I think DBI's exec will return that id, but I'm not sure.
<9> infi: thanks for that explanation makes sense now
<0> narada, np
<10> ok... INSERT INTO [...] ON DUPLICATE KEY UPDATE
<10> since I'm using PHP, how do I know after this if the query resulted in an INSERT or in an UPDATE?
<0> do you care?
<0> the data was stored.
<10> I care, since I need to present different messages to the user
<5> does anyone know a place where to find worms
<0> then I'd suggest doing it with a failed insert, and retrying the update in a second query
<0> bladhasd: your garden? a bait shop?
<10> ... which is precisely the kind of thing I was trying to avoid :/ . I was looking for a "direct" way to it
<0> bladhasd, actually, jussec
<0> I pasted a comment on the insert page a couple of years ago that talks about this.
<5> aha
<0> http://dev.mysql.com/doc/refman/5.0/en/insert.html
<0> see the third comment down
<10> hmmm interesting :)
<10> let's check that
<0> the link to the bug I posted contains more info, as well
<0> iirc, it was working as designed, and the explanation was in the bug comments
<10> it works =)
<10> thanks a lot infi
<0> cheers
<0> I struggled with the same thing, so I'm glad my hours spent frustrated helped someone else out ;)
<11> can anyone explain to me why adding the line 'ndbcluster' in my.cnf would cause mysql to report as dead on debian?
<10> infi I just noted something... I may be thinking wrong, but let's suppose that the query returns 4. is that 2 updates or 4 inserts? :)
<11> even though it actually starts
<12> dp: MySQL probably can, in its logfile. Are you actually trying to use clustering?
<11> mendel: yes
<12> Then go by the error in the logs
<11> well
<11> here's the problem
<11> the error in logs says that mysql didn't start. but a quick ps | grep mysql shows that is has
<11> but it's still spitting back the error
<12> No, the error *that mysql logged*.
<0> morphine, how many insert statements did you run?
<0> there's enough information to determine how many did inserts and how many did updates, but not which did which.
<0> that's why I think I determined to do them one at a time in the end.
<0> thus you always get 1 or 2.
<11> mendel: Feb 9 15:17:05 jackson /etc/init.d/mysql[4768]: 0 processes alive and '/usr/bin/mysqladmin --defaults-file=/etc/mysql/debian.cnf ping' resulted in Feb 9 15:17:05 jackson /etc/init.d/mysql[4768]: ^G/usr/bin/mysqladmin: connect to server at 'localhost' failed
<11> Feb 9 15:17:05 jackson /etc/init.d/mysql[4768]: error: 'Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)'
<11> Feb 9 15:17:05 jackson /etc/init.d/mysql[4768]: Check that mysqld is running and that the socket: '/var/run/mysqld/mysqld.sock' exists! Feb 9 15:17:05 jackson /etc/init.d/mysql[4768]:
<11> Feb 9 15:17:07 jackson mysqld[4670]: 060209 15:17:07 [Note] /usr/sbin/mysqld: ready for connections. Feb 9 15:17:07 jackson mysqld[4670]: Version: '5.0.18-Debian_7' socket: '/var/run/mysqld/mysqld.sock' port: 3306 Debian Etch distribution
<11> ****
<11> sorry about that
<11> http://hashmysql.org/paste/viewentry.php?id=1433
<11> there's the error
<12> Those first log entries are clearly from "/etc/init.d/mysql"
<11> but mysql is actually running, as you can see from the last 2 lines in the log
<10> well in my situation I know that it's just one row, so it's 1 or 2, but I was thinking that eventually that might not prove true
<12> !perror 2
<13> System error: 2 = No such file or directory
<12> So /etc/init.d/mysql us running /usr/bin/mysqladmin, which is complaining that it can't find the socket.
<11> correct
<11> it actually loops and tried 7 times
<12> Well, there's your problem. Diagnose that on your system!
<11> ...
<11> I've been trying to
<12> Is the socket present?
<11> at that point? not that I can tell. immediately afterwards? yes
<12> That's useful too, then. What happens if you make that 'mysqladmin' command happen a few seconds later, with a "sleep 3" before it?
<11> hold
<12> Perhaps MySQL is just taking longer to start up.
<14> www.kardelen.gen.tr
<10> wtf?
<14> www.kardelen.gen.tr
<15> hey. how would i go about copying a table (fields AND data)?
<14> www.kardelen.gen.tr
<11> /kick dosttt
<12> See, all of this is a lot more useful to us than saying that "ndbcluster" makes mysql report as dead.
<11> mendel: yes, it reports as alive after the sleep 3
<12> That's all it was, then
<11> but
<12> Adding 'ndbcluster' made mysql take a bit longer to accept connections.
<12> but?
<11> hold a sec
<11> nm then
<11> I just made it wait longer in the loop
<11> thanks for the suggestion
<16> FBH^, CREATE TABLE tablename LIKE Othertable INSERT INTO tablename SELECT * FROM OtherTable I dunno
<12> np. if it's your OS's start script, they probably just didn't consider that that'd happen
<9> gah! memory engine doesn't support TEXT type :(
<9> ERROR 1163 (42000): The used table type doesn't support BLOB/TEXT columns
<15> ty ungy
<16> FBH^, it worked?
<15> i actually just did CREATE TABLE blah_test SELECT * FROM blah;
<15> it worked perfectly, except that i had to re-add the PRI KEY
<15> no biggie.
<16> ahh ok well that could work as well
<17> is there a way to remove all text after a space in a field?
<18> !man rtrim
<13> (String Functions) : http://dev.mysql.com/doc/mysql/en/String_functions.html
<17> hmm
<17> right.. but i don't have console access, so it will have to be a phpmyadmin query
<19> moin
<16> mindamp, you can type anythign you want into a phpmyadmin query window
<18> mindame then just edit if its a one off
<19> would someone plz be so nice and can give me an hint in clustering mysql?
<17> hmm..
<17> i used this ..SELECT LEFT(login, LOCATE(' ',login)-1) AS login
<17> FROM pmd1043_users;
<17> to no avail... it displayed the new column, but it didn't take...
<19> i got 2 nodes and when i do an "create database test" it doesnt create the database in cluster mode
<16> !man Clusters
<13> Nothing found.
<16> !man Cluster
<13> (MySQL Cluster) : http://dev.mysql.com/doc/mysql/en/NDBCluster.html
<20> holas
<19> Ungy: thx
<17> anyone?
<9> guys; i have a very odd problem; if i use TEXT with memory engine it says it doesn't accept it but if i use VARCHAR then it says too big use TEXT instead
<9> "ERROR 1074 (42000): Column length too big for column 'data' (max = 21845); use BLOB or TEXT instead"
<16> Narada nice sounds liek a bug
<9> :(
<18> what was the error when using text
<9> archivist: "ERROR 1163 (42000): The used table type doesn't support BLOB/TEXT columns"
<21> hello.. i've got a number of mysql dbs at the site where i'm working these days, and i'm wondering if there's a way to specify database specific options in ~/.my.cnf. for example, can i create a [dbname] heading, with, say, host options under it? how do most people manage this sort of thing?
<16> Narada, I think you have been caught between a bug and a hard place I believe that table type doesn't support TEXT so the otehr message is incorrect when it tell you to use text or blob it shoudl tel you to use a different Engine
<9> gah :(
<18> not a bug read the doc its documented
<16> cudgel, I just type crap in
Return to
#mysql
or
Go to some related
logs:
#perl
ubuntu mount privelages
nullmailer-inject example
#math
#linux
system32.cfg error
#perl
SSH-account gentoo
mdadm reinstall linux
Virtual Printer centos
|
|