I run a SaaS service that has been growing fast and we're no longer comfortable with our existing infrastructure.

Currently, we run everything off a single linode. Even just a few months ago there were quiet hours overnight where we could do a server reboot. Now, we can't even install updates without impacting our users.

What I'm still undecided on, is what exactly our infratsurcture design should be.

I suppose the cheapest solution would be to add 1 additional Linode in a different region, and configure a MySQL master/slave relationship? Cheapest isn't the most important thing, but it has SOME importance. I believe this design would require manual failover at the DNS level.

Or, we can use a NodeBalancer with an additional linode within the same region, and again use a Master/Slave configuration?

Or?

I'm a software engineer, and I know my way around linux but I'm no ops genius, so advice is appreciated.

You can get a failover IP and have that move between nodes, that way you don't need to deal with DNS which is not a reliable failover mechanism.

_________________
Paid support
How to ask for help
1. Give details of your problem
2. Post any errors
3. Post relevant logs.
4. Don't hide details i.e. your domain, it just makes things harder
5. Be polite or you'll be eaten by a grue