> "1) is the master/slave set up described in the article going to make a duplicate copy of all the data, or does it have nothing to do with copying the database?"

The slaves are copies of the master. Every change in the master is reflected in the slaves.

> "2) will it result in higher performance somehow?"

By distributing requests across two+ droplets, you're decreasing the load on each droplet and also allowing more connections to your dns servers due to the larger amount of available resources.

I recommend setting up at least two DNS servers in two different regions so if there's an issue with one server, the second server would be unaffected and available.

But how does the slave keep a copy of the master if the database on each server isn't linked?

@m73mitchell,

Take a 2nd look at the title of the article you cited: "How to Setup DNS Slave...." The configuration outlined in that article only copies DNS records from the master server onto the slave server(s).

Thus, it will not "keep a copy of the master if (sic) the database on each server."

Thanks Pablo but then why would I want to do it? Why do I want to copy DNS records from the master server onto the slave server? What benefit does that provide?

You should have at least one slave DNS server for redundancy so it can be used instead of the other one if it's down.

Setting it up as a slave server takes care of copying the DNS records so you don't have to set up a completely separate ns2 DNS server and add each DNS record twice, once on each server.

...but then why would I want to do it? Why do I want to copy DNS records from the master server onto the slave server?

Because even the mighty GoDaddy* (*substitute whatever DNS provider you're using, here) has outages from time-to-time.

What benefit does that provide?

Here's a decent article on the topic: Considering DNS Server Design