Mysql Replication not working with WordPress XMLRPC

I have wordpress site using 3 mysql replication. Master-master replication and 1 slave.
Now have problem with the slave. When Master update posting from XMLRPC, the slave is not getting update too. This is happen a few days ago, before that is working fine and beside that at other server that using Master-Master replication is working fine too.

I can’t find any error message as If I create new post or update it not from XMLRPC, the replication in slave is working fine.

Read More

I don’t know what happen before got this problem but as long as I know, the last thing that I did was update wordpress to version 3.4. But if this the cause why at master-master replication still working fine as all site using version 3.4 too?

By the way, I saw at master there were a lot of transfer data from slave, compare from other slave server that using master/slave. About 288MB incoming data from slave.

2136K  288M ACCEPT     tcp  --  *      *       xx.14.xx.107       0.0.0.0/0           tcp dpt:3306
272K   14M ACCEPT     tcp  --  *      *       xx.18.xx.36         0.0.0.0/0           tcp dpt:3306

edit:

I just solve my own problem 😀

My configuration are 3 server.
A and B are master-master replication and C is slave-master with A.
My problem is because from XMLRPC server (D) resolve my domain to server B when create/update posting (I’m using single domain) so It’s seem can’t update slave C if is not from A.
I’m forget that I was moving to new server D too.

I don’t know if it’s normal that Slave C only can be updated from A if using XMLRPC but at least solve my problem. 😀

Related posts

Leave a Reply

2 comments

  1. Looking at your “solution” I can tell you something is wrong with your replication

    If A and B are true master-master, then writes done to B replicate to A. If C is a slave to A then all writes on A also replicate to C.

    I suspect you have your server “server-id” settings in mysql are wrong. I.E. C and B have the same server-id.

    It may be that your auto_increment_increment isn’t set to 2 (the minimum you need with 2 write masters), or that your auto_increment_offset isn’t different on each of the write masters. I suspect though if it were either of these your replication would be broke any time you wrote to both of the servers and you’d already see that problem.