List:Replication« Previous MessageNext Message »
From:api984 Date:September 24 2009 3:24pm
Subject:Re: borked replication
View as plain text  
It should only display last replication error.
you should check mysql.err or the general error log. To see whats up in
detail.

slave could get these tables if you do a dump, drop table. execute batch of
tables
that are dumped. so the master should send once again these tables. best way
is to.
stop master and slave.
copy files from master to slave.
start master
start slave

when tables are missing this is the best solution to keep sync.
remember that insert,update,delete go to master and not slave. use slave for
selects off-course.

you can use maatkit to sync, or another 3rd party program when sync is lost.
-the best way is just to copy and restart

check your config file once again to see what did you set-up to replicate.
-ignore tables in replication that you may not need.

replicate-wild? replicate-do-db?

ps. be careful when setting max-allowed-packet on master? default its 1M i
think.
-higher values caused me a lot of pain and replication got broken most of
the time
on large queries and LOAD DATA....

optimize your query's. use explain or maatkit-query-tool

ps. you are working on linux or win32 platform?

thats for now.

On Wed, Sep 23, 2009 at 6:52 PM, Andrei F <frunzales@stripped> wrote:

> Hello,
>
> SHOW SLAVE STATUS \G
>
> doesn't spit out any errors, replication is running fine at the moment.
> My concern is that some tables are not synched at all.
> For example the master server has the following tables:
>
> dbname.tablename[1,2,3]
>
> Due to the broken replication, the slave doesn't have these tables.
> SKIP_COUNTER
> won't help here since I will have to do it numerous times. I'll keep
> investigating and
> repost once I get it up and running.
>
> Regards,
>



-- 
Andrej Pintar
email : api984@stripped
          andrej@stripped
          api984@stripped
web: http://www.api984.net
contact cell: 00385 98 790 639
home server: http://api984.ath.cx

Thread
borked replicationAndrei F22 Sep
  • Re: borked replicationMiguel Angel Nieto22 Sep
  • Re: borked replicationMarcus Bointon23 Sep
    • Re: borked replicationPankaj Joshi23 Sep
  • Re: borked replicationIan Simpson23 Sep
  • Re: borked replicationAndrej Pintar23 Sep
    • Re: borked replicationAndrei F23 Sep
      • Re: borked replicationapi98424 Sep
      • Re: borked replicationMats Kindahl25 Sep