What you declare for objects
The required field is the bucket name. A prefix route can publish that bucket, and cdn_enabled on the route caches it. A retention policy can name a prefix and a retention period. Seeds can load an initial set of keys after deploy.
Storage is 0.04 credits per GB-hour. Egress is 9.00 credits per GiB. The bucket uses the project credit balance. Callers do not hold a second pair of storage-account keys in the application repository.
What stays behind in S3
An S3 bucket configuration is not copied. ACLs, bucket policies, Block Public Access settings, replication rules, lifecycle rules, and inventory reports do not become the KERNSTACK entry.
You choose the name and, when you need them, retention and a volume snapshot for a worker. You do not paste an S3 JSON policy and expect those statements to apply. Public reads go through a gateway route you reviewed.
Amazon S3 vs KERNSTACK on European Bare Metal
| Dimension | Amazon S3 | KERNSTACK Buckets |
|---|---|---|
| Operator | AWS account | KERNSTACK project |
| Manifest | Amazon S3 console or API | resources.buckets |
| Region | AWS region list | European bare metal |
| Bill | Amazon S3 line item | One project credit balance |
Frequently asked questions
No. You name the bucket under resources.buckets. Policies, ACLs, and replication rules are not copied into stack.json.
With a presigned URL for that bucket. The application does not embed a long-lived storage key.
Storage is 0.04 credits per GB-hour. Egress is 9.00 credits per GiB, on the same balance as compute.
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.