Untangling one app from its add-ons
A Heroku exit is mostly an inventory: list the add-ons, and each one maps to either a KERNSTACK resource or an honest keep. Data exports from Postgres, files from whichever storage add-on holds them, scheduled jobs become workflows, and the web process becomes functions or a worker depending on its shape.
The dyno formation does not travel, and neither do buildpacks. What travels is the code and the data; the running shape is re-declared in stack.json, which leaves the new system fully described in one file instead of a dashboard.
What the first month feels like
The add-on dashboards close one by one. The bill becomes a single credit balance with the project’s name on it, and changes — including ones an agent drafts in a sandbox — go through a plan a person reads before production sees them.
See also our detailed KERNSTACK vs Heroku comparison and the Software Lifecycle Cloud architecture pillar.
Why teams switch from Heroku to KERNSTACK
| Dimension | Heroku | KERNSTACK (Software Lifecycle Cloud) |
|---|---|---|
| Deploy model | Git push to a dyno formation | A declared project: stack.json, plan, promote |
| Everything beyond code | Add-ons rented from a marketplace, each billed separately | Tables, buckets, search, and workflows inside the project |
| Background work | Worker dynos and scheduler add-ons | Container workers and workflows with human steps |
| Bill | Dynos plus every add-on’s own plan and invoice line | One credit balance; compute at 0.01 credits per GiB-second |
| Idle cost | Dynos bill for every hour they exist | No compute charge while idle; storage only |
| Operator | Salesforce’s platform on hyperscaler regions | Our own bare metal in Europe, one zone per project |
Frequently asked questions
No — the similarity is the calm, not the model. Heroku rents you dynos and a marketplace; KERNSTACK declares a project graph — compute, data, files, search, workflows — in one file, meters it from one balance, and lets agents work on it under human approval.
That is the recommended shape. Small Heroku apps move in an afternoon: code becomes a worker or functions, data exports into tables, and the add-on accounts close behind the traffic. The big migration lessons arrive cheaply at that size.
Keep it. KERNSTACK projects call external services over the network like anything else. The goal is not purity; it is that the core of the system — code, data, jobs, search — stops being a portfolio of rented attachments.
Related capabilities, architectures, and comparisons
Ship your stack on the Software Lifecycle Cloud.
Closed preview on dedicated European infrastructure. Bring a repository or start from a blueprint.