POST /v1/sites/{siteId}/retry-provisioning

身份验证

请将 API 密钥作为 bearer 令牌发送。该密钥必须具有 sites.view 权限；缺少该权限的密钥将被拒绝并返回 403 状态码，而非 404。

此端点不需要组织 ID。您的密钥已用于识别其所属的组织，且响应范围也仅限于该组织。

免费试用

将尖括号中的内容替换为您自己的值，并将键占位符替换为您仪表板中的一个键。

curl -X POST https://api.zinndigital.com/v1/sites/{siteId}/retry-provisioning \
  -H "Authorization: Bearer zdk_live_…"

已登录？您仪表板中的 API 控制台会自动填入您真实的组织 ID 和您自己的密钥，并针对实时 API 运行请求，以便您查看实际的响应。 在 API 控制台中打开此端点

详细信息

Re-runs provisioning for a site whose build failed or never started. The transition `failed -> provisioning` has always been legal and the status enum's own comment says *"retry or delete"* — this is the door. **Idempotent, and pressing it twice is a success.** A site that is already `provisioning` answers `202` with `restarted: false` rather than a `409`: a customer pressing a button twice has not made an error. Three independent mechanisms make that safe — the workflow id is derived from the site id, the start policy is `ALLOW_DUPLICATE_FAILED_ONLY` so a completed provision can never be re-run, and the status transition locks and re-reads the row. ⛔ Nothing here calls Temporal (§2.9). The status change and a `site.created` event commit in one transaction and the existing consumer starts the workflow, so a Temporal outage cannot leave the row half-moved. A site that is `active`, `deleting` or `deleted` is a `422` — a retry re-runs the build, and those are past it. Requires `sites.view` and `hosting.deploy.manage`.

参数

响应

此端点可能返回的错误

401 · 403 · 404 · 422 · 429