Plan Compute Cap Exceeded

plan-compute-cap-exceeded · 402

The change would put more RAM or vCPUs awake at once than the account's plan allows, so the API made no change.

unit is ram-gb or vcpu, capValue is the plan's cap, currentTotal is what is awake now and projectedTotal what the change would need. The cap counts only what is awake, and gates growth only: what already runs keeps running.

Besides the members every error carries, this type adds planKey, unit, capValue, currentTotal and projectedTotal.

Common causes

  • A first deploy, a start, a wake or a new service would start compute past the cap.
  • A bigger size or more replicas would grow what is awake past the cap.

How to fix the call

  • Free room first: put a service to sleep, stop one, or choose a smaller size. Each frees its share at once.
  • Or move to a plan with a higher cap: ghaym cloud billing set-plan --plan <key>. A plan change is the user's decision.
  • Tell the user the numbers, from capValue, currentTotal and projectedTotal.

Example response

json
{
  "type": "plan-compute-cap-exceeded",
  "title": "Plan Compute Cap Exceeded",
  "detail": "The solo plan allows 8 GB of RAM across everything awake at once; this change would need 9, and 7 is awake already. Put a service to sleep, stop one, choose a smaller size, or move to a plan with more headroom with PUT /v1/billing/account/plan.",
  "planKey": "solo",
  "unit": "ram-gb",
  "capValue": 8,
  "currentTotal": 7,
  "projectedTotal": 9,
  "requestId": "req-1785231000000-4f7b2c9ae",
  "timestamp": "2026-07-28T09:30:00.000Z",
  "path": "/v1/services/0199e0a3-6f21-7c4a-9d18-2b7c5f3ad401/deployments",
  "method": "POST",
  "instance": "POST /v1/services/0199e0a3-6f21-7c4a-9d18-2b7c5f3ad401/deployments",
  "status": 402,
  "docs": "https://ghaym.me/docs/errors/plan-compute-cap-exceeded"
}