Getting account unblocked

Posted February 21, 2019 4k views
DigitalOceanBig DataUbuntu 18.04

Hey guys, new here. Been using Digital Ocean off and on for the last few years, mostly for personal projects and have really liked it a lot. Recently, though I started using it for Big Data processes. What I do is spin up the max number of nodes I can (10) and then have a script that outfits the servers, downloads the data from s3 that needs to be processed, and runs the process itself - right now doing a bunch of record linkage across millions of rows of B2B data.

Looks like Digital Ocean has closed my account now though, saying I’ve been using the servers for crypto mining which is against the TOS. Besides not being able to find anything about crypto mining in the TOS, I have never used any of the servers I’ve spun up for anything but big data analytics. Has anyone run across this before? What’s the best route to getting my account unlocked.

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
1 answer

Hey friend,

I’d like to explain a bit about the reason for this, and the thoughts behind it. Please know that I’m about to say a lot of things that may not be relevant to you. It isn’t necessarily that crypto is against our terms of service in itself, it’s that hogging shared resources for extended periods of time on our standard droplets is considered to be abusive behavior.

You see, we can’t rightly put in our terms of service every way that someone can abuse server resources. This is a constantly moving target, and it would severely confuse customers who would then be required to continually check the list for updates. Instead, it makes more sense that we focus on what it is that makes the things that would be on that list bad. What it really would come down to would be just what I said above, hogging shared resources for extensive periods of time.

A side note to that is we do specifically forbid coin mining on some of our promotions, and we state that up front. If you ever find yourself on a promotion for $100/60 days, that promotion does come with that extra stipulation.

Given that we allow customers to spin up resources and then pay for them at the end of each month, it isn’t entirely uncommon to see someone pop in with a stolen credit card, spin up the maximum amount of droplets, blast shared resources, and then skip out on the bill. Obviously our business strategy opens us up to things like that, and it’s our burden to prevent (which we work very hard to do).

Now, I’m not at all accusing you of any of that. I recognize this post as one that is going to be read many times by people other than yourself, and I know that my reply answers a lot of questions that are going to be asked in comments over the next X years. For that reason, I have expanded well beyond what is relevant to you specifically, and I have sought to answer many other questions ahead of time.

With all of that out of the way, I’d like to move to your specific scenario. When we say that you’re CPU mining and shut you down, it’s because your usage looks identical to that. Admittedly there may be some very rare situations in which a different kind of activity looks like CPU mining. We haven’t run into this enough to justify stopping to ask the question first, to be frank. Hogging shared CPU degrades performance for other customers, so the absolutely vital thing that we do immediately when we identify this is to shutdown the user who is causing the issue. There may be multiple other customers who are unable to perform at a reasonable baseline due to this.

If you do need to blast the CPU for long periods of time, we have a path for this. We have dedicated CPU threads available with our Optimized Droplets. You can take a look at those here on the Compute tab:

If you can promise me that you’ll move this workload over to our Optimized Droplets, I’ll override the previous decision and unlock your account. I’ll also toss a credit your way for your trouble.


  • Hey Jarland,

    I really appreciate your well-worded response and honestly it totally makes sense. I was not aware of the restrictions and apologize for any abuse of Digital Ocean’s services. I realize my workload is unusual and as a data guy, I totally get how that could be flagged as mining, being that CPU usage was maxed for long periods of time. That being said, I am definitely willing to switch the workload over to the Optimized Droplets - to be perfectly honest, I was going to use those originally but decided to save a couple bucks by using the Standard Droplets and maxing the CPU, unaware that that could negatively affect other users’ experiences/performance.

    I will move this and future workloads over to the Optimized Droplets, however I had a large amount of already processed data on the Standard Droplets that were shutdown. Is that salvageable or have those VMs been destroyed/deleted?

    Also in the interest of better understanding the best practices/choices for Droplet sizing and usage, are there any guidelines for average CPU/resource usage. This would help me - and other customers, I’m sure - make informed decisions which keep my costs to a minimum while respecting other users of the platform.

    Thank you so much for your help in getting this resolved.

    • Thank you for getting back to me. I absolutely appreciate your situation and I thank you for appreciating mine as well. I’ve unlocked your account and sent some credit your way to bury the hatchet. Your droplets do still exist, they were just powered off.

      It’s hard to give an exact number on resource consumption because it can, to some degree, be relative. Being a bit of an industry insider myself, when I came to DO I was actually pretty amazed at how little overcommit actually happens here. It’s kind of the best thing we don’t talk about. However, CPU is overcommitted on our standard droplets, and I can’t really publicize the exact number. It’s always subject to change based on the CPU in our latest deployments, industry trends, etc. Almost no one, relatively, ever finds out what too much is. That isn’t to say you can’t find people, because “almost no one” is still a handful of people in a really large group. It’s just not something most people need to concern themselves with.

      Thanks again for working with us, and for choosing us for your work. Here’s to better times :)

      • Thank you, Jarland. I’m back in to my account and rocking and rolling :).

        That’s interesting and as a long-time personal/commercial user of DO, AWS, GCP and Azure I’ve honestly never had a problem until now. And that’s only because I’m using hacky bash scripts to parallelize a very large big data job across a bunch of servers and maxing them out ;) - in hindsight, it doesn’t sound like a very good idea. I should really just be using actual clustering/big data technology like Hadoop or Spark but I guess tinkering with Linux servers and hacking up my own solution is more fun but I digress.

        In any case, thanks for the info and the help. Looking forward to getting back to work with DO’s services.

      • Sorry to bother you again, but it looks like I need to get my account approved or upgraded in order to use the CPU-Optimized Droplets. How do I go about getting that?