Background operations

Some operations start work in the cloud and answer 202 straight away. The work isn't finished at that point, so read the resource until it reaches a final state. Wait a few seconds between reads.

Create a project

POST /orgs/{org_id}/projects answers 202, with the project's address in Location and its default region pending.

{
  "data": {
    "id": "a9d27e10-64b3-4f18-8c05-71e3f6ab20d9",
    "name": "web",
    "description": "Storefront",
    "status": "active",
    "created_at": "2026-09-14T15:04:05Z",
    "regions": [ { "region": "yul1", "status": "pending" } ]
  }
}

Read the project until the region's status is ready or failed:

while :; do
  status=$(curl -s "$BASE/orgs/$ORG_ID/projects/$PROJECT_ID" \
    -H "Authorization: Bearer $BREQWATR_API_KEY" | jq -r '.data.regions[0].status')
  [ "$status" != "pending" ] && break
  sleep 5
done
echo "$status"

The default region is listed first.

Provision a project in another region

A project is provisioned in each extra region when you first ask for it. Find region ids with GET /regions, then:

curl -s -X POST "$BASE/orgs/$ORG_ID/projects/$PROJECT_ID/regions/$REGION_ID/provisioning" \
  -H "Authorization: Bearer $BREQWATR_API_KEY"

Response	Meaning
202, pending	Provisioning has started, or a failed run is being retried. Read the project's regions until this one is ready or failed
200, ready	Already provisioned, or the default region. Nothing was queued
409 project_not_ready	The project is still being created. Wait for its default region first

To retry after failed, send the same request again.

Delete a project

DELETE /orgs/{org_id}/projects/{id} answers 202 with the project deleting. Teardown runs in the background, and the project answers 404 once it's gone.

A project still holding resources isn't deleted. It answers 409 deletion_blocked, saying what's in the way:

{
  "code": "deletion_blocked",
  "status": 409,
  "blocker": "volumes",
  "count": 2,
  "region": "yul1"
}

Delete those through OpenStack, then try again. If a region can't be checked, the request answers 503 cloud_unavailable with Retry-After, and nothing has changed.