Auto Scaling - possible with DO? And what are your suggestions on what/how?

June 26, 2014 7.3k views

I have a centos droplet here on DO, 8GB, with a Wordpress site on it.

I'm doing about 10,000 hits a day now. The server has handled spikes up to 16,000/day with no trouble.

But were publishing more and more and I'm concerned about getting a sudden surge of traffic. I want to be able to handle the digg effect as best as I can with little to no downtime.

So I'm looking at auto scaling. I see some hosts offer it through a control panel, but DO doesn't seem to.

But I did find this:

It makes it look like I could scale up RAM, which is what I think would need during a crush... I think. Maybe CPU too? (Anyway to figure that out ahead of time?)

But I don't see anything related to the pricing for scaling up specific resources like this (I thought all DO offered was the set plans, didn't realize individual resources were available like this).

Or... would I be better off setting it up to scale up another droplet and load balancing (which I know little about)? And is that instant, or does it take a few minutes to scale up?

hmmm. I guess what I need is some suggestions on the best way to handle this issue. I'm not sure what the best solution is or if DO even has a good solution.

Appreciate any thoughts/ideas on this -


5 Answers

You can use the control panel to scale up your RAM and CPU using the "Fast Resize" option. To use it, you'll need to power down your droplet and navigate to Resize in the DigitalOcean control panel. Then select the new size you want to use. There is only about one minute of downtime. No changes are made to disk size. This uses the same pricing as creating a new droplet with that amount of RAM.

If you are considering scripting an automated solution, I'd encourage you to take a look at version 2 of the API which was just released as a public beta. It should be easier to work with in a number of way, and using it will ensure that your scripts will continue to work in the future.

If you're interested in setting up Wordpress with multiple frontend app servers behind a load balancer, we have an article on it that should point you in the right direction. It's on Ubuntu rather than CentOS, but it should give you a good idea of how you could set something like that up.

by Mitchell Anicas
A tutorial on how to use HAProxy as a Layer 4 Load Balancer for WordPress Application Servers. It assumes that your setup includes a WordPress application server that connects to a separate MySQL database server.

Thank you, I'll take a look at both those pages.

Yeh, in a crush of traffic, a minute sounds like a long downtime, that could be a hundred dollars of lost revenue in that one minute....


  • Right. In a case like that, it probably makes more sense to design an architecture that lets you just add an additional frontend Wordpress instance when needed.

    You might also look into adding a Varnish cache, if you're not doing something like that already.

Hey Chris,

So basically it looks like what DO is telling us is that there is not turn-key scaling solution that they offer. What it sounds like you'll want is to have a system that detects your load (or whatever criterion you decide to set) and scale out vertically or horizontally. And have that happen automatically.

A company called Scalr has a really nice product that I like that does exactly that. However Scalr only works with certain computing clouds, and DO isn't one of them:

But scalr is probably exactly along the lines of what you want. I've also heard good things about RightScale as well. But I haven't tried it.

So if autoscaling is what you'll need you'll probably be better off researching the pricing options of some of those other cloud vendors and give them a try.

you can simply use DO API.
we have some example configs here:

the idea is to check the load on master droplet and start new droplet if load is too high.
configure it, rsync files and inject ip into load balancer.

Hello Chris,

We support autoscaling for Digital Ocean.

How to deploy autoscaling on DO:

Also refer to:

Please let us know how we can help.


ps: Inject ip into load balancer is WIP, which can be performed by ActOnCloud Agent on HA Proxy.

Have another answer? Share your knowledge.