[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"blog-strapi-article:zh:mobile-app-network-egress-compatibility-testing":3,"blog-article:\u002Fblog\u002Fmobile-app-network-egress-compatibility-testing":625,"blog-related:\u002Fblog\u002Fmobile-app-network-egress-compatibility-testing":626,"blog-translations:\u002Fblog\u002Fmobile-app-network-egress-compatibility-testing":627},{"available":4,"article":5,"relatedArticles":137,"translations":540},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":136},"strapi",3797,"lw08jq304om4xm0qdbam6996","zh","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","Global Proxies Team",[],false,0,2,"blocks",[24,30,34,39,43,47,51,55,60,64,68,72,76,80,84,88,92,96,100,104,108,112,116,120,124,128,132],{"type":25,"children":26},"paragraph",[27],{"text":28,"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。","text",{"type":25,"children":31},[32],{"text":33,"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":21,"children":36},"heading",[37],{"text":38,"type":29},"先把 App 请求的域名列全",{"type":25,"children":40},[41],{"text":42,"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":44},[45],{"text":46,"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":21,"children":48},[49],{"text":50,"type":29},"按请求阶段记录失败",{"type":25,"children":52},[53],{"text":54,"type":29},"请求阶段不同，检查方式也不同：",{"type":56,"children":57},"code",[58],{"text":59,"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":61},[62],{"text":63,"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":65},[66],{"text":67,"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":21,"children":69},[70],{"text":71,"type":29},"对照直连与地区出口",{"type":25,"children":73},[74],{"text":75,"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":77},[78],{"text":79,"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":21,"children":81},[82],{"text":83,"type":29},"判断出口相关性与待查项",{"type":25,"children":85},[86],{"text":87,"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":89},[90],{"text":91,"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":93},[94],{"text":95,"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":97},[98],{"text":99,"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":21,"children":101},[102],{"text":103,"type":29},"常见问题",{"type":25,"children":105},[106],{"text":107,"type":29},"**Android 和 iOS 的网络栈差异会影响结论吗？**",{"type":25,"children":109},[110],{"text":111,"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":113},[114],{"text":115,"type":29},"**DNS 解析失败和连接超时怎么区分？**",{"type":25,"children":117},[118],{"text":119,"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":121},[122],{"text":123,"type":29},"**启用了证书固定（pinning），还能测试吗？**",{"type":25,"children":125},[126],{"text":127,"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":129},[130],{"text":131,"type":29},"**只在某个地区连不上，一定是代理 IP 的问题吗？**",{"type":25,"children":133},[134],{"text":135,"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。","",[138,275,414],{"source":6,"id":139,"documentId":140,"translationKey":140,"locale":9,"slug":141,"path":142,"_path":142,"title":143,"description":144,"excerpt":144,"image":145,"publishedAt":146,"updatedAt":147,"author":17,"tags":148,"featured":19,"sortOrder":20,"readingTime":149,"bodyType":22,"bodyBlocks":150,"bodyHtml":136},3827,"zoq7ywrk09kqq6smss7lhhbb","region-error-rate-rollback","\u002Fblog\u002Fregion-error-rate-rollback","国际站按地区灰度发布时，如何观察错误率并判断是否回滚？","按发布单元归并地区，先定错误率口径、最小样本与观察窗口，再把阈值写成能执行的回滚条件，并给出一次演练可复用的记录字段与判定标准。","https:\u002F\u002Fcms.quanqiudaili.com\u002Fuploads\u002Flarge_cover_2fc0fcf55b.webp","2026-10-29T03:45:17.472Z","2026-10-29T03:45:17.440Z",[],3,[151,155,159,163,167,171,175,179,183,187,191,195,199,203,207,211,215,219,223,227,231,235,239,243,247,251,255,259,263,267,271],{"type":25,"children":152},[153],{"text":154,"type":29},"国际站的错误率不该按国家逐个设。先把共享同一套代码、机制和依赖的地区合成一个发布单元，每个单元给一条错误率基线和观察窗口，再把“连续 N 分钟超过基线几倍”写成熔断条件；核心路径完全不可用时直接回滚，不必等窗口跑完。测试和灰度时，用代理IP 把出口固定到目标地区，不同地区的错误率才有可比性。下面整理一套能照做的分组口径、阈值写法和演练记录表。",{"type":35,"level":21,"children":156},[157],{"text":158,"type":29},"先按发布单元分组，不按国家分别看",{"type":25,"children":160},[161],{"text":162,"type":29},"错误率应以发布单元为单位观察，而不是把每个国家单独当成一条结论。先把使用同一版本、配置和依赖的地区归为一个单元，再为每个单元设定基线。同一版本、同一套地区配置、同一批依赖的一组地区，才能合成一个单元。确定分组后，可对照[代理 IP 产品页](https:\u002F\u002Fwww.quanqiudaili.com\u002Fproduct\u002Fdynamic-residential)整理目标地区清单，避免监控分组与实际测试范围不一致。原因是单个小市场的样本量太小，错误率的波动大部分是噪声；而把配置不同的地区混在一起，配置问题会被平均掉，看起来反而像“偶发”。",{"type":25,"children":164},[165],{"text":166,"type":29},"分组时给每个单元标注三件事：包含哪些地区、各占多少流量、当前配置版本。配置或依赖明显不同的地区单独拆出来，配置差异明显时应拆分单元，避免不同表现被汇总后掩盖。发布前先按风险缩减测试组合，把监控和人力留给高风险单元。出口类型也会影响基线：住宅 IP 走真实用户网络，机房 IP 延迟稳定，两类混在一张图里，波动会被误读成故障。",{"type":35,"level":21,"children":168},[169],{"text":170,"type":29},"统一错误率口径并设置最小样本量",{"type":25,"children":172},[173],{"text":174,"type":29},"没有口径的错误率，涨了也不知道该不该动。分子默认算服务端错误和超时；4xx 先按业务定义分类——用户输入错误、未登录、权限不足单独统计，不混进可用性，影响核心路径的 4xx（如地区策略拦截、回调地址错误）单列告警或并入对应服务指标。分母用有效请求总数，不用 UV：请求级错误率 = 错误请求数 ÷ 有效请求总数，同一个用户重试只会增加请求数，不会让分母变大十倍。若要观察“受影响用户比例”，分子分母都要换成用户口径（受影响用户数 ÷ 活跃用户数），不能拿错误请求数去除以 UV。重试、超时、4xx 的纳入规则要在统计前写进口径文档，否则同一张图换个算法就能得出不同结论。",{"type":25,"children":176},[177],{"text":178,"type":29},"小市场应设最小样本量。若指标涉及代理出口的连通性或延迟，可另参考[代理 IP 可用率、延迟和稳定性测试](https:\u002F\u002Fwww.quanqiudaili.com\u002Fblog\u002Fproxy-ip-availability-latency-stability-test)，不要把网络指标与应用错误率混为一谈。例如，团队可以设定单窗口最低请求量；样本不足时，将该地区标记为数据不足，结合发布单元整体指标和核心路径监控判断。具体样本量需按流量和历史波动确定。文中的百分比和倍数只是阈值写法示例，实际回滚标准应根据业务影响、历史基线和团队响应能力确定。",{"type":35,"level":21,"children":180},[181],{"text":182,"type":29},"用双窗口识别突发故障和持续恶化",{"type":25,"children":184},[185],{"text":186,"type":29},"一个窗口不够用，急跌和慢漂是两种完全不同的故障。两个窗口共用同一条基线，但判断逻辑分开。",{"type":35,"level":149,"children":188},[189],{"text":190,"type":29},"5 分钟窗口触发时，先暂停放量还是回滚",{"type":25,"children":192},[193],{"text":194,"type":29},"五分钟窗看的是“还能不能用”，用绝对阈值判断，例如核心路径成功率低于 A%。A% 由团队按业务基线定，并同时设单窗口最小请求量；样本不足时先标“数据不足”，不按超阈值处理。它胜在快、输在噪声大，所以触发后的动作不是立刻全量回滚，而是按下一节的统一模板走：达到最小样本且持续满足短窗阈值 → 暂停放量或摘除受影响流量 → 按业务预案确认后回滚；核心路径完全不可用时不必等窗口，直接走紧急回滚。",{"type":35,"level":149,"children":196},[197],{"text":198,"type":29},"1 小时窗口持续偏高时，怎样判断趋势",{"type":25,"children":200},[201],{"text":202,"type":29},"一小时窗看的是“是不是在变坏”，也可参考相对基线变化，例如 5xx 率达到基线 B 倍；A、B 两个值按历史波动和业务影响设定。慢速劣化靠趋势判断：可将连续多个窗口不收敛设为升级条件，单次越线先结合样本量与核心路径表现复核。",{"type":25,"children":204},[205],{"text":206,"type":29},"只用一个小时窗，整页挂掉要等一小时才动手；只用五分钟窗，慢速劣化永远触发不了。",{"type":35,"level":21,"children":208},[209],{"text":210,"type":29},"把阈值写成值班人员能执行的回滚条件",{"type":25,"children":212},[213],{"text":214,"type":29},"回滚条件必须是一句能执行的话，不能是含糊的临场判断。若故障表现为连接超时或 TLS 错误，可先按连接超时、TLS 握手和代理出口稳定性记录确认是否属于出口链路问题，再按发布指标决定是否回滚。推荐写成两句：“当某单元连续 5 分钟、样本量不低于最小请求数、核心路径成功率低于 A% 时，值班人按预案确认后在 10 分钟内回滚到上一版本。”“当某单元 1 小时窗口内 5xx 率持续达到基线 B 倍时同样处理；基线接近 0 时改用绝对阈值，避免倍数失效。核心路径完全不可用时可跳过确认直接紧急回滚。”",{"type":25,"children":216},[217],{"text":218,"type":29},"同时把三个角色写死：谁有权发起回滚、回滚到哪个版本号、回滚后由谁确认恢复。缺一项，真出事时就会卡在“谁来拍板”上。这里的 A%、B 倍、10 分钟都是示例值，按业务影响设定；“问题集中在单一地区且流量占比低于 5% 时只摘除该地区”同样是示例，低流量比例本身不能作为不回滚的理由。",{"type":35,"level":21,"children":220},[221],{"text":222,"type":29},"用一次演练验证告警和回滚流程",{"type":25,"children":224},[225],{"text":226,"type":29},"阈值没演练过，第一次真正用它，往往就是故障当天。在测试环境把错误率人为推高，完整走一遍告警、判断、回滚、恢复确认，记录每一步耗时。若演练耗时超过团队约定的响应目标，应检查值班链路、告警送达和操作权限，再决定是否调整流程。",{"type":25,"children":228},[229],{"text":230,"type":29},"> 表格中的阈值、用例编号和结果如无测试记录支持，均应视为演示样例；实际验收标准以项目需求、历史基线或团队约定为准。",{"type":56,"children":232},[233],{"text":234,"type":29},"| 字段 | 记录内容 | 用途 |\n| --- | --- | --- |\n| 发布单元 | 单元名、包含地区、流量占比 | 确认影响面 |\n| 触发指标 | 窗口长度、错误率、基线、倍数 | 判断是否达标 |\n| 执行动作 | 回滚或摘除地区、执行人、时间 | 还原决策链 |\n| 恢复确认 | 恢复时间、复核指标 | 证明真的恢复 |",{"type":25,"children":236},[237],{"text":238,"type":29},"演练记录可分别填写告警触发、值班确认、执行回滚和指标恢复的时间。若要展示时间线，应标为单次演练示例，不把示例耗时当作响应承诺。",{"type":35,"level":21,"children":240},[241],{"text":242,"type":29},"FAQ",{"type":35,"level":149,"children":244},[245],{"text":246,"type":29},"错误率基线多久更新一次？",{"type":25,"children":248},[249],{"text":250,"type":29},"每月或每个大版本后重算。用半年前的数字，阈值会失真。",{"type":35,"level":149,"children":252},[253],{"text":254,"type":29},"错误率回落就说明恢复了吗？",{"type":25,"children":256},[257],{"text":258,"type":29},"不一定。先看请求量回来没有，流量被摘走时错误率也会降。",{"type":35,"level":149,"children":260},[261],{"text":262,"type":29},"灰度期间两个版本怎么比？",{"type":25,"children":264},[265],{"text":266,"type":29},"同单元、同窗口，只差版本号。跨地区比没有意义。",{"type":35,"level":149,"children":268},[269],{"text":270,"type":29},"回滚之后还要做什么？",{"type":25,"children":272},[273],{"text":274,"type":29},"确认成功率回到基线，再决定是否重新放量。别在未确认时二次发布。",{"source":6,"id":276,"documentId":277,"translationKey":277,"locale":9,"slug":278,"path":279,"_path":279,"title":280,"description":281,"excerpt":281,"image":282,"publishedAt":283,"updatedAt":284,"author":17,"tags":285,"featured":19,"sortOrder":20,"readingTime":21,"bodyType":22,"bodyBlocks":286,"bodyHtml":136},3826,"efod1ujg76jb5w6n1moibokw","saas-regional-fallback-testing","\u002Fblog\u002Fsaas-regional-fallback-testing","某地区依赖服务不可用时，如何验证 SaaS 产品的降级与恢复？","先盘点依赖与核心路径，再在授权环境模拟超时、5xx 与 DNS 失败，验证降级提示、重试、恢复与告警，并区分核心与非核心功能。","https:\u002F\u002Fcms.quanqiudaili.com\u002Fuploads\u002Flarge_saas_cover_077e340989.webp","2026-10-29T03:45:08.380Z","2026-10-29T03:45:08.359Z",[],[287,291,295,299,303,307,311,315,319,323,340,344,348,352,356,360,364,368,371,375,379,382,386,390,394,398,402,406,410],{"type":25,"children":288},[289],{"text":290,"type":29},"降级测试要回答两个问题：核心功能是否仍可用，用户是否知道下一步该做什么。先标出哪些依赖挂了会影响核心路径，再在授权环境里模拟超时、5xx 和 DNS 失败。每模拟一种，逐项检查四件事：提示文案、重试次数、恢复后的状态、告警是否触发。降级是否按预期生效，最终要看发布后的地区错误率和回滚记录。有些问题只在特定地区的出口上出现，可在受控条件下切换出口复测。测试地区可用[全球代理 IP 产品页](https:\u002F\u002Fwww.quanqiudaili.com\u002Fproduct\u002Fdynamic-residential)规划；网络连接异常可参考连接超时、TLS 握手和代理出口稳定性记录。可以按依赖、故障注入、用户提示、恢复验证四层逐步检查。",{"type":35,"level":21,"children":292},[293],{"text":294,"type":29},"盘点依赖与核心业务流程",{"type":25,"children":296},[297],{"text":298,"type":29},"先把依赖列出来，按“挂了会不会阻断核心路径”分两类。登录、支付、数据读写这类属于核心路径的依赖，挂了必须有降级或明确的阻断提示；统计这类非核心能力可以静默降级；推荐、富文本编辑能不能静默处理，要看它是否承载核心流程、失败会不会丢数据，涉及草稿保存或内容丢失的应给明确提示。",{"type":25,"children":300},[301],{"text":302,"type":29},"每一类依赖再标四个值：超时时间、重试次数、降级方式、告警阈值。这四个值决定测试的预期，没有它们就只能凭感觉判断“体验好不好”。",{"type":35,"level":21,"children":304},[305],{"text":306,"type":29},"在授权环境模拟超时、5xx 和 DNS 故障",{"type":25,"children":308},[309],{"text":310,"type":29},"模拟只在自己可控的测试环境做，不动生产。三种失败方式分别对应三类故障：超时对应慢依赖，5xx 对应服务不可用，DNS 失败对应解析或网络配置问题。",{"type":25,"children":312},[313],{"text":314,"type":29},"模拟时只改一个依赖，其他依赖保持正常。同时改两个，出了问题也说不清是哪个触发的降级。模拟“某地区依赖不可用”之前，先确认这条依赖请求是浏览器发出的，还是服务端发出的：如果依赖由服务端发起，切换客户端出口根本不在那条链路上，测到的只是客户端到 SaaS 的一段。这类场景要在授权测试环境里用 mock、故障注入或受控网络策略，按目标地区或部署单元选择请求，分别注入超时、5xx、DNS 失败，并保留一组正常对照。仅换出口不等于已经注入依赖故障。",{"type":35,"level":21,"children":316},[317],{"text":318,"type":29},"检查降级提示、重试上限和失败状态",{"type":25,"children":320},[321],{"text":322,"type":29},"用户可见的部分有三项要查：",{"type":324,"format":325,"children":326},"list","unordered",[327,332,336],{"type":328,"children":329},"list-item",[330],{"text":331,"type":29},"**提示文案**：说清“哪个功能暂时不可用”，不要只写“系统繁忙”。",{"type":328,"children":333},[334],{"text":335,"type":29},"**下一步**：给一个用户能做的动作，比如稍后重试、改用备用方式或联系客服。",{"type":328,"children":337},[338],{"text":339,"type":29},"**可恢复性**：区分“稍后重试”和“暂时无法恢复”，别让用户反复点。",{"type":25,"children":341},[342],{"text":343,"type":29},"重试要测上限。若待测链路包含代理出口的请求重试，可参考代理请求的超时、重试和故障转移设计，但它不能替代 SaaS 依赖自身的重试与补偿验证。重试三次和重试三十次，对用户和后端的压力完全不同；超过上限后应该有明确的失败态，而不是一直转圈。",{"type":35,"level":21,"children":345},[346],{"text":347,"type":29},"验证服务恢复、数据状态和告警",{"type":25,"children":349},[350],{"text":351,"type":29},"撤销故障注入、确认依赖本身恢复健康，只是这一步的起点，不是结论。按下面五步走完再宣布恢复：",{"type":25,"children":353},[354],{"text":355,"type":29},"1. **开关状态**：观察降级开关、熔断器和限流是否自动解除，是否还有地区被锁在降级分支上。\n2. **新请求**：恢复后新发起的请求应走完整链路，结果与故障前基线一致。\n3. **遗留请求**：故障期间被拒绝、排队或超时的请求，要单独确认最终状态——是已失败、已补偿，还是仍在队列里。\n4. **数据核对**：对订单、扣款、消息投递、任务状态逐项对账，只对确认未完成的操作按幂等规则补做，不要把“有状态操作”整类重放一遍，那样会造出重复扣款。\n5. **告警与指标**：确认告警已解除、通知发到了正确的值班人、恢复后的错误率和延迟回到基线。",{"type":25,"children":357},[358],{"text":359,"type":29},"记录三组时间：故障开始、依赖解除、各项功能恢复完成；并写明每一步的预期状态与实际状态。没有这组记录，下次同类故障无法判断是恢复得慢，还是根本没恢复。",{"type":35,"level":21,"children":361},[362],{"text":363,"type":29},"记录用户可见结果和系统指标",{"type":25,"children":365},[366],{"text":367,"type":29},"记录以用户视角为主，工程指标为辅。同一次模拟，要有“用户看到了什么”和“系统记了什么”两份证据。",{"type":25,"children":369},[370],{"text":230,"type":29},{"type":56,"children":372},[373],{"text":374,"type":29},"| 依赖 | 失败方式 | 核心路径表现 | 期望降级 | 实际 |\n| --- | --- | --- | --- | --- |\n| 支付网关 | 超时 | 下单中断 | 提示稍后重试，订单保留 | 提示成功但未扣款 |\n| 推荐服务 | 5xx | 首页可打开 | 隐藏推荐位 | 显示空白区块 |\n| 短信服务 | DNS 失败 | 注册可继续 | 切换备用通道 | 直接报错 |",{"type":25,"children":376},[377],{"text":378,"type":29},"“期望降级”一列来自设计文档，不是来自这次的实际表现，否则缺失的降级会被当成正常。",{"type":35,"level":21,"children":380},[381],{"text":242,"type":29},{"type":35,"level":149,"children":383},[384],{"text":385,"type":29},"恢复后要不要补数据？",{"type":25,"children":387},[388],{"text":389,"type":29},"先看这类操作在故障期间是否留下了未完成记录：订单、扣款、消息投递这类有状态操作要逐条核对并做幂等补偿；纯展示类不需要。补偿规则应在设计阶段就写进方案，测试时验证它是否真的被执行。",{"type":35,"level":149,"children":391},[392],{"text":393,"type":29},"依赖恢复了，用户还在报错，先查什么？",{"type":25,"children":395},[396],{"text":397,"type":29},"先看降级状态是否解除，再查缓存是不是还挂着旧结果。",{"type":35,"level":149,"children":399},[400],{"text":401,"type":29},"模拟故障会触发真实告警吗？",{"type":25,"children":403},[404],{"text":405,"type":29},"会。先通知值班人，或改到测试告警通道里演练。",{"type":35,"level":149,"children":407},[408],{"text":409,"type":29},"只降级不报警行不行？",{"type":25,"children":411},[412],{"text":413,"type":29},"不行。静默降级最难发现，核心路径必须有告警。",{"source":6,"id":415,"documentId":416,"translationKey":416,"locale":9,"slug":417,"path":418,"_path":418,"title":419,"description":420,"excerpt":420,"image":421,"publishedAt":422,"updatedAt":423,"author":17,"tags":424,"featured":19,"sortOrder":20,"readingTime":21,"bodyType":22,"bodyBlocks":425,"bodyHtml":136},3825,"zq1gc1hohpgguxf0ok6xsymd","regional-network-vs-product-defect","\u002Fblog\u002Fregional-network-vs-product-defect","多地区测试发现问题，如何判断是网络故障还是产品缺陷？","用两组对照把问题收敛：同一地区内换出口 IP，再固定出口只看账号与站点侧的地区配置，按 DNS、TLS、超时、地区配置和产品逻辑分层收证，并给出可提交的工单字段。","https:\u002F\u002Fcms.quanqiudaili.com\u002Fuploads\u002Flarge_cover_bc5db86753.webp","2026-10-29T03:45:00.400Z","2026-10-29T03:45:00.368Z",[],[426,430,434,438,442,446,450,454,458,462,466,470,474,478,482,486,489,493,497,501,505,508,512,516,520,524,528,532,536],{"type":25,"children":427},[428],{"text":429,"type":29},"多地区测试出现异常时，不要先凭感觉归到网络或产品。先设计两组对照，再根据失败发生的阶段和复现结果判断。判断顺序是：先在同一个地区内换出口 IP，看问题是否跟着某一个出口走；再固定出口，只改账号、账单国家或站点侧的地区配置，看问题是否跟着配置走；最后才查产品逻辑。这里说的出口，通常就是代理IP 的落地地区，服务端据此判断你来自哪里——地区本身就是出口的属性，所以“同一个出口换三个地区”无法执行，能换的是账号与站点侧的地区配置。两组对照只说明问题与某个变量相关，不等于归因：还要排除缓存、第三方服务波动和测试时间差异。下面按对照设计、失败分层、复测和工单记录说明具体做法。",{"type":35,"level":21,"children":431},[432],{"text":433,"type":29},"先设计同地区换出口和跨地区对照",{"type":25,"children":435},[436],{"text":437,"type":29},"不用一开始就翻日志。先做两次对照，把范围从“某个地区”缩到“某个出口”或“某个配置”：固定账号、构建版本和测试数据，第一轮只换出口，第二轮在不同地区各取近时段出口，并记录地区配置。两轮结果不一致只能说明问题与某个变量相关；还需排除缓存、第三方服务波动和测试时间差异。",{"type":25,"children":439},[440],{"text":441,"type":29},"对照要成对做，不要今天换出口、明天换地区。间隔太久，版本、数据甚至第三方服务状态都变了，结论不可用。",{"type":35,"level":21,"children":443},[444],{"text":445,"type":29},"按 DNS、TLS、HTTP 和业务阶段分层收证",{"type":25,"children":447},[448],{"text":449,"type":29},"记录失败发生在哪一层，比记录“打不开”有用得多。遇到问题时可先按代理 IP 故障排查方法区分连接、目标站和本机环境，再逐项收集证据。常见五层：DNS 解析失败、TLS 握手失败、连接超时、HTTP 4xx\u002F5xx、业务层报错。层与层之间的含义完全不同：DNS 和 TLS 失败优先看出口和网络路径，4xx 优先看地区配置和权限，业务层报错才轮到产品逻辑。",{"type":25,"children":451},[452],{"text":453,"type":29},"每一层对应一个可观察字段：解析到的 IP、证书域名与有效期、连接耗时、状态码、业务错误码。拿到这五个值，研发不需要再问你第二遍。同一个现象在机房 IP 和住宅出口上表现不同：403 可能来自 WAF、权限、地区策略或应用逻辑，需结合响应头、请求 ID 和服务端日志判断，不能仅凭状态码归为网络问题。",{"type":35,"level":21,"children":455},[456],{"text":457,"type":29},"用第三次复测排除偶发问题",{"type":25,"children":459},[460],{"text":461,"type":29},"只出现一次的问题不能定性。用同一账号、同一版本、同一数据在第三个地区复测：仍然复现，按缺陷提交；不复现，标“暂未复现／原因未明”，保留前两次记录并扩大复测；只有出口对照也指向网络路径时，结论才写“偏网络”。“至少两次”可作为团队内部的示例门槛，不是通用标准；影响数据安全或核心流程的问题应立即升级，不必等待达到重复次数。",{"type":25,"children":463},[464],{"text":465,"type":29},"偶发问题也要留痕。写下时间戳、出口地区、失败阶段和当时的状态码，下一次复现时才有东西可对。",{"type":35,"level":21,"children":467},[468],{"text":469,"type":29},"写清归因结论和工单复现字段",{"type":35,"level":149,"children":471},[472],{"text":473,"type":29},"根据对照证据区分产品、配置和网络",{"type":25,"children":475},[476],{"text":477,"type":29},"偏产品、偏配置、偏网络只是当前归因方向；根因未清时写“待定位”或“疑似”，并附上证据。每种结论后面跟一句可核对的对照句，例如“同一地区换三个出口 IP 均复现，且换回原出口仍复现，可排除单一出口的网络问题”，或“账号账单国家由 A 改为 B 后价格随之变化，产品文档未声明该规则”。证据句只能说明相关；要写成结论，还需附上已排除缓存、第三方波动与时间差异的记录。",{"type":35,"level":149,"children":479},[480],{"text":481,"type":29},"记录研发复现所需的环境和请求信息",{"type":25,"children":483},[484],{"text":485,"type":29},"出口地区、时间戳、构建版本、请求 ID、失败阶段、预期与实际、复现次数、脱敏日志。缺任何一项，研发都会回头找你补，等于把一次往返变成两次。",{"type":25,"children":487},[488],{"text":230,"type":29},{"type":56,"children":490},[491],{"text":492,"type":29},"| 现象 | 对照组结果 | 更可能的原因 | 下一步 | 复现次数 |\n| --- | --- | --- | --- | --- |\n| 页面白屏 | 同一地区换 2 个出口不复现；换到第二地区（使用该地区出口）复现 | 地区相关，待定位 | 查该地区配置、缓存与共同依赖 | 2\u002F3 |\n| 请求超时 | 同一地区换 2 个出口均复现；换到另一地区仍复现 | 出口路径或共同网络依赖 | 换第二家供应商或类型交叉验证，取服务端日志 | 3\u002F3 |\n| 登录 403 | 两组对照均复现 | 产品逻辑、权限或地区策略 | 提交缺陷，附请求 ID 与响应头 | 3\u002F3 |",{"type":35,"level":21,"children":494},[495],{"text":496,"type":29},"证据不足时标注待定位或待确认",{"type":25,"children":498},[499],{"text":500,"type":29},"证据不足照样能提交，但要写清级别。实际表现偏离已确认的预期，就可以登记缺陷，根因标“待定位”；只看到一次且预期不清楚的，记“待确认”或“观察项”；涉及数据安全或核心流程不可用的，立即升级。根因查清之前，允许把“疑似”与“待定位”并列写出验证依据。",{"type":25,"children":502},[503],{"text":504,"type":29},"标注的价值在于不误导。把网络波动写成产品缺陷，团队会花两天查一个不存在的 bug；把真缺陷写成网络问题，用户会先受影响。",{"type":35,"level":21,"children":506},[507],{"text":242,"type":29},{"type":35,"level":149,"children":509},[510],{"text":511,"type":29},"机房 IP 被 403，算产品缺陷吗？",{"type":25,"children":513},[514],{"text":515,"type":29},"单凭 403 不能定性。先核对响应头、WAF 或权限日志，再在条件受控时做对照；若应用侧也能稳定复现，再交由产品团队判断。",{"type":35,"level":149,"children":517},[518],{"text":519,"type":29},"只在高峰期复现，还算地区问题吗？",{"type":25,"children":521},[522],{"text":523,"type":29},"先按时段对照高峰与平峰，再看换出口是否仍复现。换出口仍复现，只说明问题不局限于单一出口；是容量、共同依赖还是应用逻辑，还要看服务端指标和日志，不能直接归为网络拥塞。",{"type":35,"level":149,"children":525},[526],{"text":527,"type":29},"用浏览器语言能代替出口地区吗？",{"type":25,"children":529},[530],{"text":531,"type":29},"不能。语言只改界面文案，出口地区决定服务端按哪个地区响应。",{"type":35,"level":149,"children":533},[534],{"text":535,"type":29},"日志脱敏到什么程度？",{"type":25,"children":537},[538],{"text":539,"type":29},"去掉账号、Cookie 和请求体里的个人信息，保留请求 ID、时间戳和状态码。若报错来自代理 API，可按代理网关返回的错误码与自身调用日志核对。",[541],{"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":542,"featured":19,"sortOrder":20,"readingTime":21,"bodyType":22,"bodyBlocks":543,"bodyHtml":136},[],[544,547,550,553,556,559,562,565,568,571,574,577,580,583,586,589,592,595,598,601,604,607,610,613,616,619,622],{"type":25,"children":545},[546],{"text":28,"type":29},{"type":25,"children":548},[549],{"text":33,"type":29},{"type":35,"level":21,"children":551},[552],{"text":38,"type":29},{"type":25,"children":554},[555],{"text":42,"type":29},{"type":25,"children":557},[558],{"text":46,"type":29},{"type":35,"level":21,"children":560},[561],{"text":50,"type":29},{"type":25,"children":563},[564],{"text":54,"type":29},{"type":56,"children":566},[567],{"text":59,"type":29},{"type":25,"children":569},[570],{"text":63,"type":29},{"type":25,"children":572},[573],{"text":67,"type":29},{"type":35,"level":21,"children":575},[576],{"text":71,"type":29},{"type":25,"children":578},[579],{"text":75,"type":29},{"type":25,"children":581},[582],{"text":79,"type":29},{"type":35,"level":21,"children":584},[585],{"text":83,"type":29},{"type":25,"children":587},[588],{"text":87,"type":29},{"type":25,"children":590},[591],{"text":91,"type":29},{"type":25,"children":593},[594],{"text":95,"type":29},{"type":25,"children":596},[597],{"text":99,"type":29},{"type":35,"level":21,"children":599},[600],{"text":103,"type":29},{"type":25,"children":602},[603],{"text":107,"type":29},{"type":25,"children":605},[606],{"text":111,"type":29},{"type":25,"children":608},[609],{"text":115,"type":29},{"type":25,"children":611},[612],{"text":119,"type":29},{"type":25,"children":614},[615],{"text":123,"type":29},{"type":25,"children":617},[618],{"text":127,"type":29},{"type":25,"children":620},[621],{"text":131,"type":29},{"type":25,"children":623},[624],{"text":135,"type":29},null,[],[]]