All error codes

storage_content_unavailable

HTTP 503 Storage contents are temporarily unreadable

What happened

Storage contents could not be read right now. For a server this is its disk configuration; you also get this when a resize request omits cores or memory_mb and the current value cannot be read to fill in the missing one. Listing images (GET /v1/isos) answers it when the storage of the location cannot be read, and rebuilding a server answers it when the image store or the size of the server disk cannot be read - the image may well still exist, and the server was not touched. Adding a disk answers it when the disks of the server cannot be read: the data disk limit cannot be checked, so no disk is ordered. Resizing or deleting a disk answers it when the snapshots of that disk cannot be checked, and so does growing the system disk with disk_gb in PATCH /v1/servers/{id}: a disk held by a snapshot must not change, so the disk is left as it was and the detail says so. Growing the system disk also answers it, without a detail, when the disks of the server cannot be read.

What to do

Retry in a few seconds. On a resize you can avoid the read entirely by sending both cores and memory_mb. A rebuild, a disk addition, a disk resize (disk_gb of the server included) or a disk deletion refused this way has not started: repeat it as it is, the same idempotency key included.

Retrying

The request itself is fine. Repeat it later. After a 5xx to a write request, read the resource first and repeat with a new idempotency key - the operation may have started - unless this page or the detail of the response says nothing was started or changed; then, as after a 409 or a 429, repeat the request as it is.

How it arrives

{
    "type": "https://docs.fiberax.com/errors/storage_content_unavailable",
    "title": "Storage contents are temporarily unreadable",
    "status": 503,
    "code": "storage_content_unavailable",
    "request_id": "a277dcd1ca652c65"
}

Branch your code on code, never on the wording of title.