OpenAI-Compatible Image Generation API: Single-Request Recovery for Logistics Backends

Описание и применение

Готовый контракт для генерации изображений в каталоге, использующий единый OpenAI‑совместимый шлюз, обеспечивает идемпотентность и восстановление при отказах, упрощая работу с несколькими поставщиками.

Задача

Catalog image generation must be idempotent and recoverable across multiple backends without each worker knowing each provider’s SDK.

Результат

The workflow guarantees idempotent image generation, unified error handling, and provider discovery, reducing worker complexity and improving catalog reliability.

Рабочий процесс

• 1. Persist catalog revision and image variant to generate a stable job key.
• 2. Store job metadata (model, prompt, size, n) beside the key.
• 3. Send request to gateway with same shape and Idempotency‑Key.
• 4. On timeout or transient failure, retry up to 3 times with exponential backoff, honoring Retry‑After for 429.
• 5. If provider unavailable, choose another model while preserving original spec and record the change for review.

Доказательства

treat catalog image generation as an idempotent job, then put a single OpenAI-compatible POST /v1/images/generations boundary in front of the backends.

Источник

Store a stable job key derived from the SKU, catalog revision, and image variant before submission.

Источник

Reuse the same Idempotency-Key when a timeout or transient server failure makes the outcome uncertain.

Источник