[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"blog-strapi-article:zh:ios-app-regional-network-compatibility-testing":3,"blog-article:\u002Fblog\u002Fios-app-regional-network-compatibility-testing":508,"blog-related:\u002Fblog\u002Fios-app-regional-network-compatibility-testing":509,"blog-translations:\u002Fblog\u002Fios-app-regional-network-compatibility-testing":510},{"available":4,"article":5,"relatedArticles":81,"translations":465},true,{"source":6,"id":7,"documentId":8,"translationKey":8,"locale":9,"slug":10,"path":11,"_path":11,"title":12,"description":13,"excerpt":13,"image":14,"publishedAt":15,"updatedAt":16,"author":17,"tags":18,"featured":19,"sortOrder":20,"readingTime":21,"bodyType":22,"bodyBlocks":23,"bodyHtml":80},"strapi",3790,"bmafmp9ndu8mvrszo95dmvon","zh","ios-app-regional-network-compatibility-testing","\u002Fblog\u002Fios-app-regional-network-compatibility-testing","iOS App 地区网络测试：检查项与证据清单","iOS App 在特定地区请求失败时，先固定系统版本、网络扩展与权限状态，再对比直连和地区出口下的同一请求，按解析、握手、鉴权、超时记录失败阶段，最后整理成可复现的缺陷单。","https:\u002F\u002Fcms.quanqiudaili.com\u002Fuploads\u002Flarge_ios_app_cover_cd21c8afd6.webp","2026-10-28T01:59:26.066Z","2026-10-28T01:59:26.049Z","Global Proxies Team",[],false,0,1,"blocks",[24,30,34,40,44,48,52,56,60,64,68,72,76],{"type":25,"children":26},"paragraph",[27],{"text":28,"type":29},"iOS App 在某地区请求失败时，先确认设备、系统和网络配置是否一致，再比较不同出口。各项状态适用范围不同：本地网络权限仅影响局域网访问，后台状态主要影响后台任务；公共 API 失败不能仅凭本地网络权限被拒来归因。","text",{"type":25,"children":31},[32],{"text":33,"type":29},"测试时固定构建、请求和设备状态，只改变出口；记录失败发生在解析、TLS、鉴权还是响应阶段。本文聚焦自有或获授权 App 的请求测试，不讨论系统级代理配置。",{"type":35,"level":36,"children":37},"heading",2,[38],{"text":39,"type":29},"先固定设备、系统版本与网络配置",{"type":25,"children":41},[42],{"text":43,"type":29},"先记录设备型号与 iOS 版本、网络扩展状态、相关权限以及前后台状态。它们的适用范围不同：本地网络权限只影响访问局域网设备，蜂窝数据权限只影响实际走蜂窝网络的请求；后台刷新状态只用于解释后台任务。公共 API 请求失败不能仅凭本地网络权限被拒来归因。",{"type":35,"level":36,"children":45},[46],{"text":47,"type":29},"对比直连与地区出口下的同一请求",{"type":25,"children":49},[50],{"text":51,"type":29},"确认测试配置后，固定构建号、接口、请求体、账号、语言与系统状态，只改变出口。代理参数可参考[代理设置帮助](https:\u002F\u002Fhelp.quanqiudaili.com\u002Fdocs\u002Fproxy-Settings)。同一时段重复几次并保留时间戳、请求 ID 和出口记录，单次成功或失败都不足以说明稳定性。",{"type":25,"children":53},[54],{"text":55,"type":29},"安排在相近时段测试，并记录时间戳、请求 ID 与出口。要比较扩展开关时，另建一组对照，保持出口不变。",{"type":35,"level":36,"children":57},[58],{"text":59,"type":29},"按请求阶段定位失败点",{"type":25,"children":61},[62],{"text":63,"type":29},"出现失败时，记录解析结果与错误码、TLS 证书链和系统时间、HTTP 状态码，以及超时发生在建连还是响应传输阶段。401\u002F403 表示收到了应用或中间层响应，需再确认响应来源。可从授权调试构建日志或系统控制台取证，并把扩展、权限和前后台状态一起记下。",{"type":35,"level":36,"children":65},[66],{"text":67,"type":29},"记录权限、后台状态与可复现证据",{"type":25,"children":69},[70],{"text":71,"type":29},"如果只在后台或低电量状态下失败，先与前台、直连结果比较，确认系统是否挂起任务。记录设备与构建、网络配置、出口、失败阶段、错误码和复现步骤；证据不足时注明还缺哪项对照或日志。",{"type":35,"level":36,"children":73},[74],{"text":75,"type":29},"常见判断",{"type":25,"children":77},[78],{"text":79,"type":29},"测试时语言、地区、权限和扩展按本轮要验证的变量决定。若判断地区出口影响，就固定语言、权限和扩展，只改出口；若要判断扩展影响，则固定出口和其他条件，单独比较扩展开关。这样每轮结果才知道对应哪项变化。","",[82,203,369],{"source":6,"id":83,"documentId":84,"translationKey":84,"locale":9,"slug":85,"path":86,"_path":86,"title":87,"description":88,"excerpt":88,"image":89,"publishedAt":90,"updatedAt":91,"author":17,"tags":92,"featured":19,"sortOrder":20,"readingTime":36,"bodyType":22,"bodyBlocks":93,"bodyHtml":80},3797,"lw08jq304om4xm0qdbam6996","mobile-app-network-egress-compatibility-testing","\u002Fblog\u002Fmobile-app-network-egress-compatibility-testing","移动 App 网络兼容测试：五个失败阶段定位法","同一个 App 构建，直连正常、换到某地区出口就失败，问题可能落在域名解析、连接建立、证书校验、鉴权或超时任一环。按阶段记录域名、信号和错误码，才能分清兼容缺陷与出口质量。","https:\u002F\u002Fcms.quanqiudaili.com\u002Fuploads\u002Flarge_app_cover_bab910e122.webp","2026-10-28T02:01:27.065Z","2026-10-28T02:01:27.051Z",[],[94,98,102,106,110,114,118,122,127,131,135,139,143,147,151,155,159,163,167,171,175,179,183,187,191,195,199],{"type":25,"children":95},[96],{"text":97,"type":29},"同一个 App 构建，直连时正常，换到某个地区的出口就失败，这种问题最容易被一句\"网络差\"带过去。实际上失败点可能落在五处：域名没解析出来、TCP 连接没建立、TLS 握手被拒、鉴权请求被拦、或者响应超时。",{"type":25,"children":99},[100],{"text":101,"type":29},"对自有或明确授权的 App，先固定构建与设备，再列出实际请求域名。代理出口用于控制请求发起地区，不会替代真实用户网络或服务端数据。",{"type":35,"level":36,"children":103},[104],{"text":105,"type":29},"先把 App 请求的域名列全",{"type":25,"children":107},[108],{"text":109,"type":29},"列出首屏接口、图片 CDN 和第三方 SDK 等实际请求域名，并标明哪些影响核心功能、哪些可以降级。这样可避免主接口正常、某个 SDK 不可达却被误判为整条链路正常。",{"type":25,"children":111},[112],{"text":113,"type":29},"以域名清单作为测试范围，逐项记录请求阶段和结果。",{"type":35,"level":36,"children":115},[116],{"text":117,"type":29},"按请求阶段记录失败",{"type":25,"children":119},[120],{"text":121,"type":29},"请求阶段不同，检查方式也不同：",{"type":123,"children":124},"code",[125],{"text":126,"type":29},"| 阶段 | 可观测信号 | 记录方式 |\n| --- | --- | --- |\n| 域名解析 | NXDOMAIN、解析超时、返回地址与预期不符等 | 记录错误码、解析器、返回地址与耗时；内网地址也可能是合法配置 |\n| TCP 连接 | SYN 无响应、建连超时 | 记录目标端口与建连耗时，区分本地到代理、代理到目标及服务端监听\u002F路由 |\n| TLS 握手 | 证书校验失败、握手重置或超时 | 记录证书链、服务身份、系统时间与握手日志；单一信号不定根因 |\n| 鉴权\u002F拒绝 | 返回 401 或 403 | 识别响应来自目标站、WAF 还是中间层，并记录错误码与响应特征 |\n| 超时 | 解析、建连、TLS 或传输阶段超时 | 记录失败阶段与耗时分布；重复结果用于确认现象，不单独证明根因 |",{"type":25,"children":128},[129],{"text":130,"type":29},"单一信号可能对应多种原因。例如解析失败可能来自解析策略或目标站可达性；TLS 失败可能与客户端校验或链路有关。先记录信号，再结合对照组定位。",{"type":25,"children":132},[133],{"text":134,"type":29},"记录请求停在哪个阶段，以及错误码和耗时。再用相同构建与请求条件比较直连和地区出口；差异只说明相关性，仍需结合服务端或供应商证据。",{"type":35,"level":36,"children":136},[137],{"text":138,"type":29},"对照直连与地区出口",{"type":25,"children":140},[141],{"text":142,"type":29},"对照时固定构建、设备、系统、语言、账号和请求参数，只改变出口。重复运行以观察结果是否稳定，并记录执行时段。",{"type":25,"children":144},[145],{"text":146,"type":29},"记录时间戳、出口地区和实际检测结果。先确认直连基线稳定；否则加入地区出口后难以区分问题来源。",{"type":35,"level":36,"children":148},[149],{"text":150,"type":29},"判断出口相关性与待查项",{"type":25,"children":152},[153],{"text":154,"type":29},"多次结果可帮助判断现象是否稳定，但单一出口失败只能说明与该出口条件相关。要判断责任层，还需补同地区其他出口、不同时间和服务端证据。",{"type":25,"children":156},[157],{"text":158,"type":29},"TLS 失败可能涉及证书链、系统时间、目标站策略或中间链路；多个出口结果随机波动也不能单凭现象确定为网络质量问题。应保留证据继续核对。",{"type":25,"children":160},[161],{"text":162,"type":29},"工单写清已验证的条件、失败阶段和仍缺的证据；证据不足时标为“待确认”，再决定由客户端、服务端或网络人员继续排查。",{"type":25,"children":164},[165],{"text":166,"type":29},"记录时可用简表列出请求域名、失败阶段、直连与地区出口的实际结果、错误码\u002F耗时及仍缺的证据。示例域名和结果不代表真实测试数据。",{"type":35,"level":36,"children":168},[169],{"text":170,"type":29},"常见问题",{"type":25,"children":172},[173],{"text":174,"type":29},"**Android 和 iOS 的网络栈差异会影响结论吗？**",{"type":25,"children":176},[177],{"text":178,"type":29},"会。两端的解析、连接复用和证书校验实现可能不同，应分别执行相同用例。",{"type":25,"children":180},[181],{"text":182,"type":29},"**DNS 解析失败和连接超时怎么区分？**",{"type":25,"children":184},[185],{"text":186,"type":29},"DNS 失败要核对错误码、解析器和策略；TCP 超时还要区分本地到代理、代理到目标及服务端监听\u002F路由。TLS 失败需核证书链、服务身份、系统时间与握手日志；401\u002F403 要识别由目标、WAF 还是中间层返回。每种现象都记录阶段和证据，不仅凭单一信号归因。",{"type":25,"children":188},[189],{"text":190,"type":29},"**启用了证书固定（pinning），还能测试吗？**",{"type":25,"children":192},[193],{"text":194,"type":29},"能做，但要看固定的是哪一层。只固定叶子证书的实现，换出口后证书不变，通常仍能建立连接；固定中间证书或整条链的实现，遇到地区链路上的中间设备改写就容易失败，这本身也是要记录的现象。这里不提供绕过证书校验的方法，测试应在授权的调试构建上进行。",{"type":25,"children":196},[197],{"text":198,"type":29},"**只在某个地区连不上，一定是代理 IP 的问题吗？**",{"type":25,"children":200},[201],{"text":202,"type":29},"单一出口失败不能直接说明是代理问题或 App 缺陷。结合失败阶段，补同地区多出口、供应商、时段、直连与服务端证据；DNS、TLS、401\u002F403 和超时都应先核实具体响应来源与路径，再决定是否换出口复测。",{"source":6,"id":204,"documentId":205,"translationKey":205,"locale":9,"slug":206,"path":207,"_path":207,"title":208,"description":209,"excerpt":209,"image":210,"publishedAt":211,"updatedAt":212,"author":17,"tags":213,"featured":19,"sortOrder":20,"readingTime":36,"bodyType":22,"bodyBlocks":214,"bodyHtml":80},3796,"y64gki0k633wq4rop2v89b94","website-third-party-script-regional-fallback-testing","\u002Fblog\u002Fwebsite-third-party-script-regional-fallback-testing","国际站第三方脚本测试：降级与告警验证","第三方脚本在部分地区加载失败时，如何判断页面降级是否符合设计？本文说明如何盘点页面依赖、在自有测试环境模拟阻断与超时，并检查核心流程、恢复行为和告警。","https:\u002F\u002Fcms.quanqiudaili.com\u002Fuploads\u002Flarge_cover_af12243bd0.webp","2026-10-28T02:01:10.949Z","2026-10-28T02:01:10.933Z",[],[215,219,223,227,231,235,252,256,260,274,278,282,286,290,294,298,302,306,310,314,318,322,326,330,334,337,341,345,349,353,357,361,365],{"type":25,"children":216},[217],{"text":218,"type":29},"第三方脚本在部分地区被阻断或超时，页面往往\"看起来还是正常的\"：布局没塌、图片还在，但表单提交不了、埋点全丢、支付卡在最后一步。这类第三方资源加载失败通常不抛错，因此比白屏更难发现，也更难排查。",{"type":25,"children":220},[221],{"text":222,"type":29},"按依赖影响安排测试：先检查阻断核心流程的脚本，再看非核心组件。地区出口用于观察资源是否可达；降级逻辑则在自有测试环境中通过故障注入验证。",{"type":35,"level":36,"children":224},[225],{"text":226,"type":29},"盘点页面依赖",{"type":25,"children":228},[229],{"text":230,"type":29},"先按页面列出第三方脚本、来源域名、用途和失败影响。同一脚本在不同页面可能有不同优先级；例如验证码脚本可能阻断注册流程，在内容页却只是增强功能。记录来源域名，便于把单个资源不可达与整页故障区分开。",{"type":25,"children":232},[233],{"text":234,"type":29},"可按影响分为三类，具体等级应由产品路径确认：",{"type":236,"format":237,"children":238},"list","unordered",[239,244,248],{"type":240,"children":241},"list-item",[242],{"text":243,"type":29},"阻断：核心动作做不了，比如登录、支付、提交表单，先测。",{"type":240,"children":245},[246],{"text":247,"type":29},"降级：功能变弱但可完成，比如推荐位不显示，其次测。",{"type":240,"children":249},[250],{"text":251,"type":29},"无感：埋点、AB 实验这类，最后测。",{"type":35,"level":36,"children":253},[254],{"text":255,"type":29},"在授权环境模拟常见失败",{"type":25,"children":257},[258],{"text":259,"type":29},"在自有测试环境模拟失败，不针对生产流量或第三方真实服务。可分别测试：",{"type":236,"format":237,"children":261},[262,266,270],{"type":240,"children":263},[264],{"text":265,"type":29},"阻断：在自有测试环境对指定请求做拦截或拒绝连接；不要改生产 DNS。",{"type":240,"children":267},[268],{"text":269,"type":29},"超时：让请求挂起不返回，观察页面的等待上限。",{"type":240,"children":271},[272],{"text":273,"type":29},"错误响应：返回 5xx 或格式错误的脚本内容，看是否抛异常并中断后续脚本。",{"type":25,"children":275},[276],{"text":277,"type":29},"分开测试后，才能看出页面对网络阻断、等待超时和错误响应的处理差异。",{"type":25,"children":279},[280],{"text":281,"type":29},"用出口去复现\"某些地区加载不了\"，拿到的是现象：代理 IP 能证明这个问题在你关注的出口上真实存在，但证明不了你的降级逻辑对不对。两件事要分开做，前者用观测，后者用模拟。若需检查代理连通性，可参考[智能检测助手](https:\u002F\u002Fhelp.quanqiudaili.com\u002Fdocs\u002FSmart-check)，再回到自有环境的故障注入验证。",{"type":35,"level":36,"children":283},[284],{"text":285,"type":29},"检查失败时页面如何响应",{"type":25,"children":287},[288],{"text":289,"type":29},"降级策略按依赖等级、已批准的安全设计与产品契约验收。核心安全依赖失败时，明确提示并受控阻断可能是正确设计；替代验证必须经安全审查且不得降低校验强度。非核心组件是否需要提示，也按产品体验要求判断。",{"type":25,"children":291},[292],{"text":293,"type":29},"**页面布局**",{"type":25,"children":295},[296],{"text":297,"type":29},"检查脚本未加载时布局是否跳动、核心控件是否被遮挡。是否构成缺陷按产品体验要求判断。",{"type":25,"children":299},[300],{"text":301,"type":29},"**核心流程与安全边界**",{"type":25,"children":303},[304],{"text":305,"type":29},"按已批准的产品方案验证主流程。验证码等安全依赖失败时，可能需要安全阻断并明确提示；不要临时接入未经审核的替代验证，也要确认页面不会误报成功或重复提交。",{"type":25,"children":307},[308],{"text":309,"type":29},"**状态提示**",{"type":25,"children":311},[312],{"text":313,"type":29},"检查失败状态是否清楚可见，避免页面无响应或误报操作成功；提示和后续操作按产品设计验收。",{"type":25,"children":315},[316],{"text":317,"type":29},"对于受影响的流程，检查页面表现、操作结果和数据状态是否一致。",{"type":35,"level":36,"children":319},[320],{"text":321,"type":29},"验证恢复和告警",{"type":25,"children":323},[324],{"text":325,"type":29},"网络恢复后，检查脚本是否重复初始化、用户行为是否重复上报，以及告警是否按配置触发。",{"type":25,"children":327},[328],{"text":329,"type":29},"告警这一项最容易漏测。日志断言只能证明应用记录了事件，不能证明告警规则触发或通知送达。已接入测试通道时，验证规则、观察窗口、通知路由、接收与恢复解除；没有通道时明确标注“告警送达未验证”。",{"type":25,"children":331},[332],{"text":333,"type":29},"记录页面、脚本、模拟方式、用户可见结果和告警状态。日志存在不代表告警送达。",{"type":35,"level":36,"children":335},[336],{"text":170,"type":29},{"type":25,"children":338},[339],{"text":340,"type":29},"**模拟失败会不会影响线上真实用户？**",{"type":25,"children":342},[343],{"text":344,"type":29},"若故障注入严格限定在自有测试环境或隔离会话，并确认规则不会影响生产流量，通常不会影响线上用户。测试前先核对拦截范围；不要对第三方真实服务主动探测或施加阻断。",{"type":25,"children":346},[347],{"text":348,"type":29},"**脚本失败但页面看起来正常，还需要处理吗？**",{"type":25,"children":350},[351],{"text":352,"type":29},"视影响等级和安全设计决定。重点确认用户是否能识别失败、核心流程是否按预期处理，以及数据有没有误报成功或重复提交。",{"type":25,"children":354},[355],{"text":356,"type":29},"**降级做到什么程度算通过？**",{"type":25,"children":358},[359],{"text":360,"type":29},"按依赖等级、安全设计和产品约定判断。受控阻断可能是正确结果；替代流程不得绕过安全校验。日志断言通过也不代表告警已经发送或送达。",{"type":25,"children":362},[363],{"text":364,"type":29},"**用代理 IP 复现脚本失败就够了吗？**",{"type":25,"children":366},[367],{"text":368,"type":29},"不够。出口只能证明\"在这个地区确实加载失败\"，证明不了降级逻辑写对了。降级要单独测，出口只是发现问题的入口。本地请求拦截或 mock 可模拟失败，但不能证明目标地区出口或第三方实际故障；地区可达性观察与本地故障注入要分开记录。",{"source":6,"id":370,"documentId":371,"translationKey":371,"locale":9,"slug":372,"path":373,"_path":373,"title":374,"description":375,"excerpt":375,"image":376,"publishedAt":377,"updatedAt":378,"author":17,"tags":379,"featured":19,"sortOrder":20,"readingTime":21,"bodyType":22,"bodyBlocks":380,"bodyHtml":80},3795,"gfk97rtcs8x8vegewhqnyala","mobile-app-push-timezone-network-delivery-testing","\u002Fblog\u002Fmobile-app-push-timezone-network-delivery-testing","App 推送时区测试：三层时间与边界用例","推送\"延迟\"常被当成网络问题，实际是触发时间、到达时间、展示时间三层语义混在一起。固定基准时钟，覆盖 UTC 边界与夏令时切换，分别记录三层时间与点击后的落地状态，才能判断问题在哪一层。","https:\u002F\u002Fcms.quanqiudaili.com\u002Fuploads\u002Flarge_app_cover_2edd2f32cb.webp","2026-10-28T02:00:54.047Z","2026-10-28T02:00:54.036Z",[],[381,385,389,393,397,401,405,409,413,417,421,425,429,433,437,441,445,449,453,457,461],{"type":25,"children":382},[383],{"text":384,"type":29},"推送延迟要先分清服务端调度、推送服务状态、设备接收和屏幕展示。用户反馈的“没收到”可能发生在其中任一环节。",{"type":25,"children":386},[387],{"text":388,"type":29},"先统一测试设备与服务端的时钟，再分别记录调度触发、服务端收到的推送状态、设备实际接收和通知展示时间。设备时区与网络出口是不同变量，应分开测试。系统推送和用户点击后加载页面也可能走不同网络路径；只有实际配置走代理的请求才使用该出口。本文聚焦自有 App 的验证。",{"type":35,"level":36,"children":390},[391],{"text":392,"type":29},"分别记录调度、接收与展示",{"type":25,"children":394},[395],{"text":396,"type":29},"这几个时间点由不同环节产生，记录时应使用同一通知 ID，并区分服务端日志与设备侧观测。",{"type":25,"children":398},[399],{"text":400,"type":29},"**调度触发时间**",{"type":25,"children":402},[403],{"text":404,"type":29},"查看计划触发时间与任务实际执行时间，并记录使用的时区。固定偏移可能来自本地时间与 UTC 转换不一致，不应先归因为地区网络。",{"type":25,"children":406},[407],{"text":408,"type":29},"**设备接收时间**",{"type":25,"children":410},[411],{"text":412,"type":29},"推送服务受理或投递状态不等于设备已经收到。优先记录设备侧通知日志；无法观察设备接收时，标为未知。通知类型、优先级、平台规则和系统状态都可能影响结果，应结合通知 ID 和统一时钟分析。",{"type":25,"children":414},[415],{"text":416,"type":29},"**屏幕展示时间**",{"type":25,"children":418},[419],{"text":420,"type":29},"检查通知实际出现的时间、顺序和分组。系统策略可能造成延迟或摘要展示；先记录现象，再按产品预期、平台规则和同类设备结果判断是否异常。字段缺失、落地错误、延迟、丢失或重复都应有记录。",{"type":25,"children":422},[423],{"text":424,"type":29},"分别比较计划触发、实际执行、服务端状态、设备接收和展示。差异能帮助定位环节，但单一时间差不能直接证明根因。",{"type":35,"level":36,"children":426},[427],{"text":428,"type":29},"测试跨日与夏令时边界",{"type":25,"children":430},[431],{"text":432,"type":29},"覆盖 UTC 跨日、跨月与跨年边界。夏令时测试使用 IANA 时区而非固定偏移量，例如 `Europe\u002FBerlin` 的春季跳时与秋季重复时段。测试前先明确不存在或重复的本地时间应跳过、顺延还是执行多次，并记录计划与实际执行时间。",{"type":25,"children":434},[435],{"text":436,"type":29},"尽量在同一批次执行对照，并记录每条用例的计划与实际触发时间。",{"type":35,"level":36,"children":438},[439],{"text":440,"type":29},"分开测试设备时区与网络出口",{"type":25,"children":442},[443],{"text":444,"type":29},"测试时先固定出口，只改设备时区，观察展示时间和日期文案；再固定时区，只改变出口，检查地区化页面或接口响应。不要把出口变化直接当作推送投递差异的原因。",{"type":35,"level":36,"children":446},[447],{"text":448,"type":29},"检查点击后的落地页面",{"type":25,"children":450},[451],{"text":452,"type":29},"点击行为应与推送投递分开验证。记录冷启动、已登录和链接过期等场景下的落地页面、耗时与回退结果，并注明当时的设备时区和实际网络路径。",{"type":25,"children":454},[455],{"text":456,"type":29},"冷启动时记录首屏耗时和目标页面是否保留；已登录时检查页面状态；链接过期时检查提示与回退页面。",{"type":25,"children":458},[459],{"text":460,"type":29},"实际测试可用一行记录计划时刻与 IANA 时区、对应 UTC、调度和设备侧证据、点击结果。春季跳时和秋季重复时段分别建用例；设备接收无法观测时写“未知”，未实际执行时不要填写通过或失败。",{"type":25,"children":462},[463],{"text":464,"type":29},"推送服务到设备的链路与 App 请求可能不同，应根据实际配置分别观察。代理出口可用于测试配置走该出口的地区化请求；目标地区测试机或云主机也可提供对照。",[466],{"source":6,"id":7,"documentId":8,"translationKey":8,"locale":9,"slug":10,"path":11,"_path":11,"title":12,"description":13,"excerpt":13,"image":14,"publishedAt":15,"updatedAt":16,"author":17,"tags":467,"featured":19,"sortOrder":20,"readingTime":21,"bodyType":22,"bodyBlocks":468,"bodyHtml":80},[],[469,472,475,478,481,484,487,490,493,496,499,502,505],{"type":25,"children":470},[471],{"text":28,"type":29},{"type":25,"children":473},[474],{"text":33,"type":29},{"type":35,"level":36,"children":476},[477],{"text":39,"type":29},{"type":25,"children":479},[480],{"text":43,"type":29},{"type":35,"level":36,"children":482},[483],{"text":47,"type":29},{"type":25,"children":485},[486],{"text":51,"type":29},{"type":25,"children":488},[489],{"text":55,"type":29},{"type":35,"level":36,"children":491},[492],{"text":59,"type":29},{"type":25,"children":494},[495],{"text":63,"type":29},{"type":35,"level":36,"children":497},[498],{"text":67,"type":29},{"type":25,"children":500},[501],{"text":71,"type":29},{"type":35,"level":36,"children":503},[504],{"text":75,"type":29},{"type":25,"children":506},[507],{"text":79,"type":29},null,[],[]]