How to ensure that in shared environments one user doesnt stall the CPU for another user. Here are some suggestions that Digital Ocean could implement to increase customer satisfaction. implement resource monitoring for droplets (i.e digital ocean should implement it), and send notifications to droplet owners if their hypervisor is coming close to a threshold before stalling the droplets of other customers (i guess the threshold needs to be passed for X number of seconds) add to usage policy that if stalling occurs in a droplet digital ocean reserves the right to pause the droplet offer automated threshold tools to droplet owners to increase resource allocation until a max by which the droplet will freeze -- this is also beneficiary to digital ocean as users buy more resources as needed, and automatically. The core problem is that droplet owners must be responsible citizens to their peers, all of whom share the cost of the shared computer infrastructure. It's not acceptable to let one take over a resource unchecked from all others. i hope this customer feedback helps, if you can pass it on to the product manager -- who will thank you by letting them know how customers feel digital ocean service offered can be significantly improved.
