WPCloudDeploy Documentation

DigitalOcean Provider: Introduction, Installation & Configuration Guide

Installing the DigitalOcean Provider

The DigitalOcean provider is built into the core plugin so no installation is necessary.

Configuring the DigitalOcean Provider

Before you can start to use DigitalOcean cloud servers you need to tell WPCD how to talk to them.

You can provide this information in the WPCLOUDDEPLOY → SETTINGS → CLOUD PROVIDERS → DigitalOcean tab.  There are several steps to this process:

  • Make sure you upload the public portion of  an SSH Keypair to your DigitalOcean dashboard.  Or you can use the CREATE SSH KEY-PAIR button to automatically generate a pair for you. Either way you should make sure there is an SSH public key in your DigitalOcean account before performing any other steps!
  • Enter your DigitalOcean API Key – you can get this key from the API section of your DigitalOcean dashboard.
  • Click the SAVE BUTTON
  • Click the SAVE BUTTON AGAIN.  This will cause the SSH keys area to show up.
  • Select one of the public SSH keys from the drop-down.
  • Enter the PRIVATE portion of the ssh key pair into the Private SSH Key field.  This private portion must be the one that is associated with the public key you selected above.
  • Enter the key password if one is used on the keypair.
  • Click the SAVE BUTTON
  • Click the SAVE BUTTON AGAIN.

At this point your DigitalOcean provider should be configured.  All other information is optional.


Recently we’ve noticed that smaller servers fail to deploy properly at DigitalOcean.  You will see one of two error messages in the “terminal” when a server fails to deploy:

  • sudo: dos2unix: command not found. – This one is usually followed by this message: Unfortunately the server provisioning process has failed. Please delete this server and restart! Sorry about that!
  • Failed – quitting process – This one shows up after attempting to add and configure additional repositories.

In general you just have to delete the server from our server list and then retry the deployment.  Sometimes choosing a different location helps and sometimes just choosing a different size will succeed.

Depending on the time of day, we might have to try 3 or 4 times to get a server to deploy all the way.

Our best guess for this is that, during deployment we are hammering at the server since we’re installing a ton of software packages.  So any virtual server that is slightly unstable will be prone to fail sooner rather than later.  Many of DigitalOcean’s data centers are close to capacity and it is possible that some new small servers are simply being provisioned on machines that are already crowded. Choosing a different size or location will generally cause the new VM to be attempted on a different physical machine.

The upside is that servers that deploy fully will be far more stable and less likely to suffer a VM failure in the future.