If there is a common method of scaling with MySQL databases it is the
Read Replication Cluster solution.

Most websites start out with a single database and grow from there.
If the site's content is being generated from their database then
they will eventually hit a wall with reads from the database. Tuning
and hardware will buy you some growth but in the end disks spin only
so quickly. Luckily most websites are predominantly read intensive
and for this reason replication will solve scaling problems for many
people. Replication is a means by which MySQL sends updates of one
database to one or more databases which will act as a slave. These
changes are atomic, which means the changes are applied in full. No
row will ever be partially updated, and no transaction will be seen
on the slave that did not commit on the master












  [Read more...]

I've never been all that interested in solving small problems. Small problems with scaling are resolved with single indexes, upgrades to hardware, or simply creating a bigger pipe.

When the measure of the Internet was a T-1, you could flood the network with the average 486. At the time I watched people buy hardware in the hundred's of thousands, and sometimes more, which never went used. Today's hardware is overkill for a lot of applications, so the first step in scaling is often tuning the hardware that you have already purchased. Make use of what you already have.

The "Slashdot Effect" is a perfect example of what is normally a small problem. What is the Slashdot Effect? Point tens of thousands of eyeballs at a website and watch it crash. The root cause of this? Most of the time it is because the site operator had their Apache max connections set to some ridiculous number.



  [Read more...]