科技 · 微软服务故障持续报道 · 第 4 篇,共 5 篇
微软将延迟邮件保护期限延长至48小时,服务故障尚未完全恢复
签证处理和医院通信凸显了事件的影响:微软将延迟邮件的保护期限延长至48小时,但仍报告服务不稳定、容量尚未完全恢复。
文字 阅读需 7 分钟 · 219 个来源
原语言视频
译自英文原文。
自上次报道以来的变化
2 小时后
当时
微软服务中断导致医学影像检查和最终轮求职面试延误,服务恢复仍不全面我们当时说尚无定论的问题
故障起因、经审计的受影响用户和机构总数、总体损失以及连接器全面恢复的时间,仍未明确。
尚无定论故障起因、经审计的受影响用户和机构总数、总体损失以及连接器全面恢复的时间,均尚未确定。新通知增加了对延迟邮件的保护,但并未解答这些悬而未决的问题。
容量恢复后,随着流量回升,服务能否持续响应;以及微软承诺发布的事件报告是否会解释为何多次恢复尝试失败。
尚无定论微软仍称服务状况不稳定,容量尚未完全恢复,因此流量回升后服务能否持续响应尚无定论。微软承诺在故障缓解后发布完整的事后报告,涵盖根本原因和保护措施,但未给出发布日期。199
现在
签证处理和医院通信凸显了事件的影响:微软将延迟邮件的保护期限延长至48小时,但仍报告服务不稳定、容量尚未完全恢复。
微软现已为这次故障中受影响的邮件增加了一项具体保护措施:将邮件超时时限延长至48小时。其11月13日10:01 UTC更新的公开事件通知称,多个工作负载正在改善,但服务状况仍不稳定,容量也尚未完全恢复正常。事件分类仍是“服务降级”,而非已恢复。199
这一区别很重要。48小时就是两天,但这既不是邮件送达承诺,也不是对故障持续时间的预测。微软表示,延长时限旨在保留延迟的邮件,同时优先处理队列和积压邮件。微软尚未说明有多少邮件正在等待、可能丢失了多少邮件,或积压何时能够清理完毕。199
在我们11月13日的报道中,我们于09:26 UTC播报了持续的服务中断,同时提到微软通过流量优化取得的初步进展。我们还报道了具体后果:医疗扫描延迟或取消、职业治疗中断、发票处理受阻,以及错过最后一轮求职面试。这些是具体工作无法进行的个案,并非经审计的经济损失,也不证明故障造成了人身伤害。
10:01的通知为恢复情况增添了明确的邮件保留措施,并再次确认容量仍低于基线水平。它并未用全面恢复的结论取代尚未完成的恢复工作。199
恢复必须经受流量回升的考验
哪些功能可用,取决于依赖关系
微软的通知列出了20项产品和服务。这并不意味着20个应用都完全无法使用。许多故障涉及与Exchange Online(微软的云端电子邮件系统)交互的功能,而不依赖该系统的功能仍可使用。这有助于解释为何同事面对同一事件时,体验可能大不相同。199
Teams尤其需要谨慎区分。微软表示,大多数核心功能——包括加入通话和会议、群聊及频道消息——不受影响。日历安排、邀请和会议修改可能失败,创建实时或虚拟活动以及某些联系人、聊天记录和Copilot操作也可能失败。标准频道仍可正常使用。直接访问SharePoint Online文件也仍然可用,不过外部共享和依赖电子邮件的功能可能受影响。199
更广泛的影响波及Assignments(作业)和Reading Progress(阅读进度)等教育功能;Copilot访问工作邮件、日历和文件的功能;Intune设备注册和合规性检查;以及Planner和To Do的操作。涉及邮件、日历和联系人的Microsoft Graph请求可能超时,导致关联应用和通知中断。199
安全和合规团队也面临各自的限制。Defender页面可能无法加载,而依赖Purview保护信号的策略评估可能延迟或失败。Purview问题包括授权、标记、数据丢失防护、审计和调查搜索、导出,以及访问受保护内容。Fabric和Power BI在处理带有敏感度标签的材料时也可能发生故障。这些是影响重大的功能受损,并非攻击者已侵入客户系统的证据。199
应对措施因应用而异。网页版访问可能对部分Power Apps和Field Service用户有所帮助;已经登录的用户或许可以通过保持登录状态,且不清除应用数据或重新安装,来避免中断。只有在组织的保护规则允许时,才可考虑移除受影响报告材料的敏感度标签——这不是要求普遍降低安全性的指示。199
邮箱之后等待处理的工作
在10:01的更新中,微软承诺于11月13日11:00 UTC前发布下一次更新。这是提供信息的截止时间,而非恢复服务的时间。真正的检验仍是容量能否持续稳定,以及积压邮件能否清理完毕:邮箱能够打开,并不能证明邮件已经送达,也不能证明那些等待处理的公共服务工作已经赶上进度。199
时间线
第 4 天 · 5 篇报道
我们的播报 · 第 4 次更新当前位置
微软将延迟邮件保护期限延长至48小时,服务故障尚未完全恢复微软所称客户受到实际影响的结束时间。
微软表示,所有仍受影响的客户和地区服务均已恢复,并继续监控延迟邮件。
初步报告发布;拉曼纳在 X 上致歉。
我们的播报 · 第 5 次更新较新
微软将全球故障追溯到自身代码中7月出现的一个漏洞,但尚未解释为何该漏洞一直未被发现。
来源163
按媒体查看所有来源
来自 163 家媒体的 219 个来源· 引用最多的优先
Google News Singapore (SG:en)channelnewsasia.com5
- 首次报道于
- Some Singapore government systems affected by Microsoft outage: GovTech - CNA
- Microsoft 365, Outlook down for thousands of users - CNA
- Microsoft outage drags on for hours, disrupting Outlook, Teams for thousands, including Singapore users - CNA
- Microsoft Outlook outage: Disruptions hit workers worldwide amid setback in recovery efforts - CNA
- Australia's spy agency, federal police hit by Microsoft outage - CNA
Workhorse Group, Inc.workhorse.com5
BBCbbc.com4
- 首次报道于
- Microsoft outage: Outlook and Teams impacted globally as company struggles to restore service - BBC News
- Outlook: falha na Microsoft deixa email fora do ar email e afeta Teams, Microsoft 365 e outras plataformas; o que se sabe
- Una falla en Microsoft deja fuera de servicio el correo electrónico de Outlook y otras plataformas
- Microsoft outage sees Teams, Outlook and Copilot suffer 'connectivity issues'
Google News Pakistan (PK:en)theguardian.com4
- 首次报道于
- Microsoft ‘restarting’ as outage impacts global users, with Outlook and Teams affected | Microsoft | The Guardian
- A global Microsoft 365 outage has hit thousands – what happened and when will it be fixed? - The Guardian
- Microsoft outage live: Microsoft scrambling for new ways to fix global outage after initial attempts fail to restore Outlook services | Microsoft | The Guardian
- ‘Looks like we’ll have to talk to people’: Australia sent back to time before email with major Microsoft outage | Microsoft | The Guardian
Google News السعودية (SA:ar)cnn.com4
- 首次报道于
- Global Microsoft outage disrupts Outlook, Teams and other services
- Global Microsoft outage leaves millions without access to Outlook email, Teams and other tools - CNN
- "عطل" يصيب مايكروسوفت 365 وأوتلوك.. والشركة تعلق - CNN Arabic
- Massive Microsoft outage disrupts businesses, schools and governments around the world | CNN Business
franceinfofranceinfo.fr3
- 首次报道于
- DIRECT. Crise sociale : la réunion entre les syndicats de fonctionnaires et le Premier ministre a débuté
- Une panne mondiale touche la messagerie Outlook de Microsoft, "aucun signe d'action malveillante ou de cyberattaque" n'a été trouvé à ce stade
- DIRECT. Panne mondiale de Microsoft : "Des résultats positifs en matière de stabilité du service" constatés par le géant américain
播出原貌
- Microsoft has extended its email timeout to forty-eight hours to protect delayed messages. Hospitals, government agencies and businesses are waiting on recovery, while Microsoft's available capacity remains below normal.
- On November thirteenth, we reported continuing disruption and Microsoft's early gains from traffic optimization. Its latest assessment describes further improvement alongside unstable service. Can Microsoft restore enough lasting capacity to clear the mail and restart the work waiting behind it?
- Reports of problems came from seventy-four countries and regions, across the Americas and parts of Europe, Asia and Africa. The disruption began Thursday and continued into Friday, reaching public services as well as workplace inboxes.
- The international reach is clear. The number of people affected hasn't been counted. Submitted outage reports measure complaints, and their totals depend on the product, place and time being measured.
- Peak reports exceeded one hundred thirty thousand in the United States. A worldwide Outlook snapshot approached one hundred forty thousand. Smaller national snapshots also showed problems. These figures cover different windows and scopes, so adding them would exaggerate what they tell us.
- Microsoft says servers unexpectedly became unavailable, pushing their work onto the infrastructure still running. The company distinguished that loss of usable capacity from a sudden surge in customer traffic. Some servers returned to service, then stopped responding again when requests came back.
- Restarting a server only helps if it keeps responding when work returns. Engineers have been managing traffic while bringing capacity back. At the same time, they have to process the email that accumulated during the disruption.
- Recovery measures have included pausing deployments, restarting servers, rolling back changes and testing another way to check server health. Microsoft also rerouted and restricted traffic. Selected regions improved, including an earlier improvement in North American Teams issues, while other customers continued having problems.
- Microsoft's latest assessment puts those two experiences together. Several services are improving, but health still varies and capacity hasn't fully recovered. Earlier failed repair attempts predated Friday's progress, so continuing disruption can coexist with genuine improvement.
- The extended timeout gives delayed email a two-day preservation window while Microsoft prioritizes queues and backlogs. It doesn't promise every message will arrive within two days, or predict an outage lasting that long. Microsoft hasn't given a time for clearing the backlog.
- Subscribe to this channel as we follow Microsoft's recovery and the public services waiting for it.
- Microsoft still hasn't confirmed what first made the servers fail. Server configuration and a faulty update have been suggested as possibilities. Australian Home Affairs Minister Tony Burke relayed the update explanation as Microsoft's apparent assessment.
- Mike Burgess, Australia's intelligence chief, and federal police commissioner Krissy Barrett confirmed their organizations were affected. Australia activated its priority-restoration agreement with Microsoft. Those confirmations concern affected systems and communications within agencies that also carry out many other operations.
- Microsoft says it has found no evidence of malicious activity associated with the incident. That's its current security assessment while the cause remains under investigation. The company plans to publish its root-cause findings and further protective measures after mitigation is complete.
- Microsoft's notice lists twenty product-and-service entries. Across those services, many failures involve features that interact with Exchange Online, its cloud email system. Other functions remain available, which helps explain why colleagues can have very different experiences of the same incident.
- Outlook users may struggle to open mailboxes, send or receive messages, or read message bodies and attachments. A sender and subject line can appear while the contents won't open. Problems can affect web and desktop Outlook, mobile synchronization and software connections to Exchange.
- Some Microsoft 365 administrative pages may fail to load. Recently used file views can break, while administrators can lose access to cloud update, policy and inventory controls. Locally stored documents and offline editing can remain usable.
- Education features that may be affected include Assignments, Reading Progress and Speaker Progress. Copilot Chat may fail to load or answer. Other Copilot experiences can lose access to work email, calendars and files needed to complete a request.
- Most core Teams functions, including joining calls and meetings, group chats and channel messages, remain unaffected. Calendar scheduling, invitations and meeting edits can fail. Creating Live or Virtual events can also fail, alongside some contact, chat-history and Copilot operations.
- Standard Teams channels remain functional. Direct access to files in SharePoint Online also remains available, although external sharing and email-dependent features can be impaired. Microsoft Graph, the software connection used by other applications, can time out on mail, calendars and contacts, affecting connected Teams operations and notifications.
- Device enrollment, check-ins and compliance checks through Intune may be intermittent. Planner has plan and task reliability problems, and To Do may fail to load or finish operations. Power Apps and Field Service mobile users may encounter sign-in failures.
- Web access may offer an alternative for Power Apps and Field Service. People already signed in may avoid disruption by staying signed in, keeping application data and avoiding reinstallation during the incident. Those options depend on the application and the user's situation.
- Security teams face problems too. Defender for Office 365 can experience delays or failures when evaluating policies using Purview protection signals. Defender XDR pages may not load. Features governing software identities can also fail when they depend on sensitivity labels.
- Purview problems reach authorization, automatic labeling and data-loss prevention. Protected content may not open. Audit searches, investigation searches and exports, classification and on-demand scans may be impaired. Communication compliance, protection policies and role-based access controls are also affected.
- Fabric and Power BI can have trouble with sensitivity-labeled files and exports. That includes PDF and PowerPoint exports, updates to labeled material and opening labeled Power BI Desktop files. Removing a label is an option only where an organization's protection rules allow it.
- Universal Print may fail to submit jobs or perform administrative operations. Windows 365 connections can be impaired for customers using Purview Customer Key protection for their tenants and Cloud PCs. These feature failures can disrupt work without showing that attackers have breached a customer.
- Immigration New Zealand announced that all visa application processing was paused until further notice. Customers also couldn't apply through its enhanced online platform. Its foreign ministry experienced trouble receiving email at home and at overseas missions, while some offices and businesses reverted to manual methods.
- Most Norwegian ministries supported by the government's digital-services organization were affected, with the Justice Ministry identified as an exception. US House users were affected too. Officials in Japan also described government-system impacts. In Australia, reported disruption extended to schools, courts and businesses.
- The University of Iowa hospital network lost email and voicemail access and directed patients to its online portal. Unity Health Toronto declared a Code Grey alert after email failed, while saying patient and resident care was unaffected and its hospitals remained safe and open.
- Earlier personal accounts showed other clinical workflows breaking down. Maria Derosa said scans after a car accident were delayed or cancelled because she and the clinic couldn't access referrals. Occupational therapist Lauren Van Ingen said she couldn't work with clients.
- Melbourne job-seeker Manny Flores said he missed a final Teams interview, rescheduled for the following week. Adelaide business owner Peter Roche said authentication problems blocked accounting and invoice processing. He described losing thousands of dollars as orders and invoices stopped moving.
- Those were individual experiences described in the earlier November thirteenth report. Students and vocational-education users also described interrupted study support and virtual classes. The accounts show specific work that couldn't proceed, without providing a total financial loss or a medical outcome from the delayed scans.
- IndiGo and CemAir reported messaging difficulties without reporting flight disruption in those announcements. An earlier second-hand account described a flight to Papua New Guinea waiting for printed pre-flight documents. The carrier, length of that delay and final outcome weren't identified.
- Air India said its flight operations remained unaffected. Those different experiences show how a communications failure can interrupt one workflow while another continues. They also explain why an operating meeting, clinic or aircraft can't stand in for every service dependent on the same platform.
- OpenAI continues investigating problems with new Microsoft connector connections and related tools. Anthropic's latest displayed update still describes most Microsoft 365 connector mail, calendar and search requests failing. Claude for Outlook remains affected in that update, which dates from Thursday night.
- Earlier Anthropic updates also described SharePoint, Teams chat and add-in problems. These failures affect assistants' access to connected Microsoft information. Workhorse Group also postponed its quarterly results and results-and-outlook call because of the outage. The call moved to November sixteenth at four-thirty in the afternoon, US Eastern.
- Microsoft has added protection for delayed messages while it rebuilds capacity. But service health remains variable, and the incident still carries the classification of service degradation. Sustained capacity and backlog clearance are the tests still ahead, with no full-restoration time announced.
- Microsoft's next update is due by eleven UTC on November thirteenth. That's a target for more information. The company has promised its full post-incident findings after mitigation, but hasn't set a publication date.
译文更新于





