Operating models
Deciding whether you operate your application yourself or we take it on.
The most important decision on gh0stcloud is not a technical one. It is: who looks after the application once it is running?
The platform underneath is identical either way. What differs is how much of the responsibility sits with us and, accordingly, what it costs.
Running it yourself
This fits when someone on your side can and wants to own deployments. Concretely: that you are comfortable with container images and application configuration, writing and changing manifests, rolling back a failed release, keeping dependencies current, and judging what more replicas or more storage will mean at the end of the month.
That is not a high bar. If you are building the application anyway, you can probably already do it.
We operate the platform. You operate your application.
Having it operated
This fits when nobody on your side should take that role, or when there is someone, but their time is worth more elsewhere.
We then take on an agreed part of running the application: deployments, configuration changes, applying updates, backups and recovery, and keeping an eye on whether it is working. What exactly is included is written down before anything starts, not out of formality, but because "you'll look after it" without a list gets read differently by each side on the day it matters.
Who answers for what
| Run it yourself | Have it operated | |
|---|---|---|
| The platform | Us | Us |
| Protection rules and account limits | Us | Us |
| Contents of the Git repository | You | Shared, per agreement |
| Application code | You | You |
| Deciding what gets released | You | You, unless explicitly handed over |
| Performing the deployment | You | Us, where agreed |
| Administering the application | You | Us, where agreed |
| Incident: platform | Us | Us |
| Incident: application | You | Us, within the agreed scope |
| Price | Lower | Higher |
How to decide
The useful question is not which model has more features. The platform features are the same. It is: what happens on your side when the application stops starting on a Friday evening?
If there is a clear answer and that person has the time, run it yourself and keep the money. If there is no answer, or the answer is "then someone calls Falk", the managed model is exactly what you are buying, with the difference that it is written down beforehand.
You can move in either direction later. Switching is a commercial change, not a technical migration.