Is there anyone out there looking for auto scaling (horizontal and vertical) solutions for their Digital Ocean projects?

Posted January 22, 2019 4.6k views

I’m working on an auto scaling platform - the MVP is complete for Ubuntu distributions (I posted some screenshots on Twitter last year).

Is this something that people still need or has everyone worked their way around this feature?


These answers are provided by our Community. If you find them useful, show some love by clicking the heart. If you run into issues leave a comment, or add your own answer to help others.

Submit an Answer
2 answers

That’s interesting. Are you able to do vertical scale down? What are the caveats?


  • Yup, scaling down is supported. Scaling is built on a rules engine that supports various metrics including RAM & CPU usage, web response times, error rates, queue length, etc.

    Not sure there are many caveats. The obvious one right now is that the queue length metric needs some work to be done by the customer (i.e. your solution needs to let our platform know how much work needs to be processed). And the another would be latency between deciding to scale down and actually dropping droplets - by the time you drop a droplet, you might need to scale the system back up.

    The latter probably means that you’ve set up your rules too aggressively though…(and we don’t support scaling up and down at the same time - it needs to happen sequentially).

    • How do you deal with DB and/or storage scaling, synchronization, replication, etc? Got a link to share?


      • Unfortunately we don’t support DB/storage scaling - this is purely a web and worker scaling solution.

        It would be interesting to tackle the former, but well outside scope :)

        Our solution is evented and uses HA message bus. Synchronization is done at our platform level as well as at the load balancer level.

We’ve just opened our beta sign up: