| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Comments:
<0> upgrade to 5.0.16-r4
<0> dump your databases with the compat flag set to 4.1
<1> 1) don't downgrade
<0> downgrae to 4.1
<0> import
<1> 2) if you didn't do it via mysqldump you can't do it
<0> then tickele Therion
<2> i did mysqldump but it produced a 0 byte file
<0> wow ok
<1> you might take that as an indication that you did something wrong
<0> might
<2> hah
<1> Why downgrade, though?
<2> well, that's why i'm here looking for help from the pro's..
<1> 5.0.19 just came out
<2> i downgraded b/c gentoo kept trying to force a downgrade on an emerge and i couldn't get it to stop, it annoyed me so i performed a manual downgrade
<2> so, should i uninstall 4.1.xx and go back to 5.0.16 to see if my databases are still there?
<2> i'm willing to try - i just don't know where to go from here
<2> i can login to mysql and see my three databases listed
<2> Therion is 5.0.19 stable?
<2> hello, anyone?
<1> It's current release in 5.0
<2> Therion but it's not marked as stable yet is it?
<1> was marked as stable at 5.0.15
<1> though I think .19 definitely deserves that more than .15
<3> is there something wrong with this statement --> event_id SMALLINT NOT NULL AUTO_INCREMENT UNSIGNED,
<1> minDscrm: UNSIGNED in wrong location
<3> it's a part of a CREATE TABLE SCRIPt
<2> hmm...., so you think if I uninstall 4.x and up to 5.x I should have my databases back and website back functioning?
<3> Therion, where does it go?
<1> smallint unsigned
<1> spudnick: Since I don't know what problem you were having, I cannot say with certainty.
<3> Therion, rad thanks
<2> Therion well, my webpage is blank, and upon entering phpmyadmin, i get a database error (can't remember and am reinstalling that just in case)
<1> damn vonage has the slowest website on the planet
<4> Any tips for moving a live database onto another machine?
<4> is there any sort of magical synchronisation thing mysql can do ?
<4> Hmm.
<2> Therion, no more issues with phpmyadmin
<2> a reinstall cured that - but - my website is still blank
<2> boggles my mind - makes it hurt - i don't like it
<2> ok, i think i may have found out the prob - phpmyadmin now reports, from mysql, that my databases were created with a different version and that they cannot be read
<2> so, off to unmerge 4.x and remerge 5.x
<2> i guess
<1> That seems unlikely
<1> what's the actual error message?
<2> #1033
<1> ugh
<2> when i clicked on my gallery database, it responded with "database was created with different version of mysql and cannot be read"
<1> 5 can read 4
<1> sounds like you are maybe running with 5's databases
<2> but not the other way around
<1> right
<2> yep, so, i'll go back to 5 and hope nothing is screwed
<5> if I am doing a subquery in mysql how do I refer to a value in the parent query select parentid,value from tableone where something = (select otherid from table2 where id = parentid);
<6> I have two table 'users' (contains 'name' and 'location') and 'locations'
<6> err, my enter key slipped
<6> actually, if I just have a table 'locations' with fields 'name' and 'location', can I return the list of all locations, but with the one corresponding to the current 'name' first in the list, all in pure SQL (vs. doing it in php)
<1> "yes"
<6> well, could you show me an example query?
<1> depends on what you meant by "the one corresponding to the curent name first in the list"
<1> all locations fo ra particular name, or just put a particular name first?
<6> a particular name first
<1> select name, location from locations order by name='currentname' desc; would be one thing
<1> where you would get all of them, with the first item in the list corresponding to name='currentname'
<6> ah, first time I see that in the order by clause, very usefull
<6> thanks, sql is always full of surprises ;)
<7> gday. im looking to do master/slave replication, but the docs i've read mention that it doesn't guarantee that the slave will be exactly in sync at any one point in time. im wondering how the sync happens - does the slave poll the server? or server push to client?
<7> err s/server/master s/client/slave/
<8> agrajag: Two processes: one copies binlogs from the master to the slave, and the other executes the copied binlogs
<8> The connection originates at the slave
<7> ok, so how often does the slave copy binlogs from the master?
<1> constantly as long as it is connected
<4> Agrajag-: slave polls server.
<7> ok, how often does it poll?
<1> ...constantly...
<4> when changes happen.
<8> Still constantly
<4> Im trying to figure out how to do this too.
<7> what, every nanosecond?
<1> Every time a change is made
<7> how does it know to poll exactly when a change is made?
<1> Heh
<1> It's connected to the master
<7> that's not the definition of poll...
<7> i realise that
<1> When a change is made ont he master, it pulls the change
<8> agrajag: A permanent connection is made from the slave to the master. The master writes binlogs to disk, and the slave reads binlogs from disk.
<4> Will mysqldump --all-databases dump all the stuff I need to replicate the database to a completely new system?
<7> mendel: how does it know when the binlog is changed?
<1> jsoft: "maybe"
<7> mendel: it must poll the binlog right?
<4> Agrajag-: read the manual man.
<7> i have!
<4> http://dev.mysql.com/doc/refman/4.1/en/replication-implementation.html?ff=nopfpls
<7> ok not that one, cheers :)
<4> Agrajag-: the server will send the client some sort of indication that there have been changes made.
<4> Geez.
<7> err that still doesn't explain it
<1> Go read the source code.
<4> Oh you read it all just like that did you?
<7> before people said that the slave polls, now you're saying the server pushes?
<1> If you're going to be that ridiculously anal about it, that's your only option.
<8> agrajag: The server writes to a file. The slave reads from a file.
<4> Agrajag-: do you have any idea what push and poll means?
<7> jsoft: no, none at all
<8> I presume the slave uses read() and seek() but I haven't looked.
<4> poll is a function. it is used to wait for data on a socket/fd
<4> so the server will notify the client when there is a change made, and the client will grab those changes.
<4> Damn it, who gives a ****, use it or not use it.
<7> jsoft: of course i know what poll and push are...
<4> Oh so you do know what they are now.
<4> Tell me, team. If I used rsync to replicate a mysql db to another new system, that would be ok, right?
<4> While its going? ( as long as there is no differences between the files )
<1> Agrajag-: slave connects, downloads whatever binlogs it doesn't have; once it's at the end it waits for the server to let it know the log is updated
<4> like, a script that rsyncs them and I keep running it till there are no differences
<1> Agrajag-: then it pulls more
<1> Agrajag-: rinse/repeat
<7> Therion: yep i've got that figured now, thanks
<1> jsoft: should rsync, then after you have the initial copy, FLUSH TABLES WITH READ LOCK; and rsync for changes
<1> jsoft: shouldn't really use rsync, IMHO, but some people do it
<1> If you use InnoDB you'll want to actually halt mysqld before the final rsync
<9> think it's ok to wait til main thread is waiting for activity
<9> as long as you keep traffic shut off
<1> flupps|US: tricky to guarantee that
<1> it might flush buffer pool or do an insert buffer merge or something
<1> might be okay if you keep rsyncing I guess, but I don't think I'd want to trust that
<4> hehe :)
<9> heikki did a post about a year ago about how to take a binary copy on a running server
<1> flupps|US: I find that hard to believe, if it didn't involve Hot Backup
<4> Yeah, I think I might do the rsync it, then shut it down before the final rsync routine. That aught to be safe.
<1> here's Heikki
<1> "Tomas,
<1> did you use rsync to copy the ibdata file of a RUNNING mysqld server? "
<1> hee
<6> an off the topic question on load balancing, is it generally better (in terms of performance), to have one web server and one mysql server, or to just do replication between the two ?
<1> What would you replicate between a web server and a database server?
<6> no, I mean each server has httpd and mysql, and replicate each other
<6> clustering
<1> Depends on particular environment
<1> Usually you have far more web servers than database servers
<1> or one of each :)
<1> They don't generally stay in 1:1 relationship as you scale
<1> as they scale differently
<6> of course
<1> !kick therion DIE COMMIE
<10> is primary_key(col1, col2, col3) the best way to ensure that the combination of the values in the 3 columns must be unique but any 1 or two can be duplicates?
<10> i'm currently using unique index.. but it doesnt seem to be working..
<6> actually, I was wondering how much more ressources it takes to make a query to a remote mysql server, rather than on localhost ?
<10> my entry looks like: http://hashmysql.org/paste/viewentry.php?id=1556
Return to
#mysql
or
Go to some related
logs:
macosx pyreadline
fc5 ntpd.conf
scanftime POSIX
reinstall liby2util*rpm
dnat 139 port not work
boot from flashdrive
kmail dcop unread mail
vim column copy
dominic_ hotmail
#bash
|
|