The problem stemmed from reaching a network software limit. We've increased this limit to prevent such issues from happening again.

The nodes are recovering, we're still investigating the cause of this downtime.

We are aware of the issues with the VPS cluster in Amsterdam and are working on resolving the issue.