To add a few short notes: 1) What happens when you modify data on the slave directly depends on how you configure your setup. It is possible to have slave updates appear on the master, that is usually referred to as circular replication. Since you have 2 replication slaves I would advise against using circular replication (1 -> 2 -> 3 -> 1). I would also highly recommend you use the 'read-only' flag on your slave to prevent accidental updates which would break data consistency.
2) In a setup like this, I would recommend that you consistently name your database handles appropriately in your code, say $dbh_write and $dbh_readonly. Atle - Flying Crocodile Inc, Unix Systems Administrator On Fri, 6 Jan 2006, Jason Williard wrote: > I am trying to understand exactly how replication works. So far, I see that > changes made on a master server are replicated to the slave server(s). > However, if a change is made on a slave server, is that replicated back to > the master as well as all other slaves? > > > I am asking this question as I try to develop a plan for more efficient web > servers. Here is what I am planning. Please let me know if this sounds > smart, or like a bad idea. > > Server 1: Redhat MySQL Master > Servers 2 & 3: Load-Balanced Redhat Apache web servers w/MySQL Slaves > > Servers 2 & 3 will be serving the same content and will need access to the > same data from the MySQL server(s). I am hoping that running MySQL on each > of the web servers will help to reduce the overall load on the servers. > > > ---- > Thank You, > Jason Williard > > > > > -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
