is 2-way the same as clustering

[datamile]

hi,

I've not figured this out from the documentation. I've a php-mysql application, that we are looking to load-balance, and fail support, by have 3 copies of the database running in different locations etc.

The for example, we will either round-robin the databases for balancing, or have a primary, and then secondary which in sync should comms timeout.

Can someone explain the 2-way update ? i.e our application is recording internet traffic details. So in a single database world you would see

insert into traffic (user,bytes) values (1,0)

update traffic

set bytes=bytes+10

where user=1

update traffic

set bytes=bytes+20

where user=1

giving bytes = 30

now in a multi database world, each statement might go to either server, but we would like each server to have the value of 30 when they are both available. i.e server 1 gets 1 and 3, server 2 gets 2…

Will server 2 update fail as it has missed the insert ?

[Shadow]

2-way sync feature of SQLyog is not a clustering solution! It was designed to bring 2 dbs in sync with one another, that's true, but completely lacks load balancing and automatic update when a change was commited to any of the dbs. What you need is MySql clustering. You'll find articles relating clusters on MySql's website here.

[Author]

Posts