{% extends "_layout.html" %} {% set title = "Download too large" %} {% set status = "403 Forbidden" %} {% set description = "An object over the API’s download ceiling. Does not refill — use the portal." %} {% block main %}
The artifact you asked for is larger than this deployment serves over the API, limits.max_download_bytes. It is there and you may have it; it is refused for its size alone. The ceiling is static, so waiting and retrying return the same answer.
size in the listing with limits.max_download_bytes in GET /v1.logs and reports artifacts may be under the ceiling where the whole node archive is not.Formerly 429 limit-exceeded with limit: "max_download_bytes" and a Retry-After. That was a retry loop: a client obeying Retry-After on a ceiling that never refills retries for ever.
{
"type": "https://sc.example.com/server-errors/download-too-large",
"title": "Download too large",
"status": 403,
"detail": "...",
"instance": "/v1/jobs/01J9...",
"limit": "max_download_bytes", // REQUIRED on this type
"trace_id": "4bf92f3577b34da6a3ce929d0e0e4736"
}
| Member | ||
|---|---|---|
limit | REQUIRED | Always max_download_bytes. |
trace_id | optional | The request’s W3C trace id. Quote it in a support request. |
detail is prose and may be reworded at any time. Branch on type, and read the members above for the specifics — never parse detail.
GET /v1/jobs/{id}/artifacts/{artifact_id} |
The object is larger than the ceiling. |
GET /v1/jobs/{id}/logs |
A finished node’s log archive, where the redirect to it would hand you an object over the ceiling. |
limit-exceeded is a ceiling that refills, and carries Retry-After.
upload-too-large is the same kind of ceiling on what you send. It is a 413 because it is the request’s own size; here the request is small and the answer is not.
artifact-not-approved is an object you may not download at all, whatever its size.