进一步提升关键站点的可用性 Raise availability for the sites that matter most
高可用规则为页面级容灾能力，仅处理页面导航请求。本页将明确说明其实际工作机制与能力边界，包括何时触发、如何探测和跳转、何时显示离线页，以及哪些请求不在覆盖范围内。该功能仅面向增强版和定制版开放。 HA rules are a page-level failover capability that handles page-navigation requests only. This page spells out how it actually works and where it stops: when it triggers, how it probes and redirects, when the offline page appears, and which requests are out of scope. Available only in the Enhanced and Custom editions.
页面级容灾，覆盖范围明确Page-level failover, with a clear scope
为关键站点启用高可用规则后，系统会注入页面容灾逻辑。当主地址不可达时，页面导航请求可切换至预设的备用地址；如无备用地址可用，还可向访客展示离线页。Once HA rules are enabled for a critical site, the system injects page-failover logic. When the primary is unreachable, page-navigation requests can switch to a preset backup address; if no backup is available, an offline page can be shown to visitors.
只有整页导航请求会进入容灾逻辑；图片、脚本等子资源请求不会被接管。Only full-page navigation requests enter the logic; sub-resource requests like images and scripts are not taken over.
页面导航请求会先按正常流程访问主地址。主地址响应成功时，响应直接返回，不触发跳转。A navigation first hits the primary as usual; when the primary responds successfully, that response is returned directly, with no redirect.
主地址请求失败后，系统会以 no-cors 方式对备用地址进行轻量探测，超时时间约为 2.5 秒。只有探测确认备用地址可访问后，才会通过 302 跳转至该地址；跳转时保留原路径、查询参数和锚点，并设置粘滞状态。After the primary request fails, the system runs a lightweight no-cors probe of the backup with a timeout of about 2.5s. Only once the probe confirms the backup is reachable does it 302-redirect there, preserving the original path, query and hash and setting a sticky state.
如果备用地址探测也失败，系统将返回可配置的离线页，避免访客直接看到原始错误信息。If probing the backups also fails, the system returns a configurable offline page instead of exposing the raw error to visitors.
把 CDN 的控制权掌握在自己手中Keep full control of your CDN
按节点数和授权站点数（以根域名 Zone 计）透明计费，可根据实际部署规模选择合适的授权档位。Transparent pricing by node count and licensed-site count (counted per root-domain Zone) — choose the licensing tier that matches your deployment.