Microsoft 365 outage: what Microsoft has actually said, in order
A sourced timeline of the November 12 Microsoft 365 and Exchange Online outage (EX1550543, MO1550821): what Microsoft said, when it said it, and what admins should and should not do.
Lots of guessing going around today (DDoS, bad patch, you name it), so here is what Microsoft has actually posted, in order. The admin center is login-only, so the quotes below come from the full Service Health text, which UPenn mirrors publicly (find EX1550543 and click Details to expand the update history). Each timestamp is the time of that admin center entry. Where a timestamp links to an @MSFT365Status post, the X post is a shortened version of the same update and usually went out a few minutes later. TechRadar's live blog reproduced several of the full entries. All times ET. Current as of Microsoft's Fri 11:15 PM update; post-incident report status added Mon Nov 16.
Update, Mon Nov 16, 1:57 PM ET: Microsoft says its post-incident report is out. At 11:47 PM ET on Sunday, Nov 15, Microsoft moved both incidents (EX1550543 and MO1550821) to "post-incident review published," with one line: "A post-incident report has been published." The report lives in the Microsoft 365 admin center, which needs a tenant login, and the public mirror I've been quoting does not carry its text. Microsoft also doesn't say whether this is the preliminary report (promised within two business days) or the final one (within five). I'm not going to summarize a document I can't link you to.
What Microsoft's public closing status for EX1550543 says now:
- Window: "Thursday, November 12, 2026, at 4:20 PM UTC" to "Friday, November 13, 2026, at 4:10 PM UTC." That is 11:20 AM ET Thursday to 11:10 AM ET Friday.
- Preliminary root cause, in Microsoft's words: "a failure in the systems responsible for establishing connections to the Exchange Online service. Specifically, traffic shifts during standard network maintenance and upgrades exposed a latent bug causing a rapidly increasing number of requests to queue for processing. As this accumulated, the affected servers exhausted their available processing capacity and could no longer accept new connections, resulting in widespread service access failures." Microsoft adds that "our analysis of the underlying root cause is ongoing."
- Still: "This was a technical issue not related to cyber or AI / agents."
Correction. On Nov 13 I wrote that Microsoft's final status put the official start at 10:05 AM UTC (5:05 AM ET). That is what the entry said when I copied it at 11:38 PM ET that night. Microsoft has since revised that same entry, keeping its original timestamp, to a 4:20 PM UTC start, which lines up with its first admin center entry at 11:20 AM ET. The Start Time field at the top of both incidents still shows 10:05 AM UTC, so Microsoft's own record disagrees with itself; I'm going with the dated statement. I also relabeled the "earlier network capacity upgrade" quote below: Microsoft still says that sentence in its final status, but its "Preliminary root cause" line now carries the text above. Earlier text is left in place with these corrections marked.
Latest (Fri 11:15 PM ET): Resolved per Microsoft at 11:15 PM ET Friday. Microsoft now shows Service restored for both EX1550543 and MO1550821. Its final status says "This was a technical issue not related to cyber or AI / agents" and describes the cause as "an issue during an earlier network capacity upgrade that surfaced following a recent network change," which it first labeled a "Preliminary root cause" (see the Nov 16 update above for the current wording). Microsoft said it would provide a preliminary Post-Incident Report within two business days and a final one within five business days.
What's affected
- EX1550543, Exchange Online, first listed with an 11:17 AM start on Nov 12 (Microsoft's final status first gave the official start as 10:05 AM UTC; it has since revised it to 4:20 PM UTC, which is 11:20 AM ET. See the Nov 16 update.): mailbox access through Outlook on the web, MAPI (Outlook desktop), Exchange ActiveSync and REST.
- MO1550821, opened 1:26 PM and also posted on Microsoft's public status page: Teams, SharePoint Online, Copilot, Defender XDR, Planner, To Do, Intune, Defender for Office 365, several Purview services (DLP, auto labelling, communication compliance, data classification), Power BI export of sensitivity-labelled reports to PDF, the M365 Apps admin center (Cloud Update, Cloud Policy, Inventory), Universal Print, Windows 365 Cloud PCs that use Purview Customer Key, some admin center pages, and Microsoft 365 Education. The list grew through the afternoon of Nov 12.
- At 7:00 PM Microsoft rewrote the MO1550821 scope and drew the line clearly: "Features of the Microsoft 365 services below that interact with Exchange Online and email services are impacted. Elements of those services that do not depend on interaction with Exchange Online remain functional." For Teams, "Core messaging and online meetings remain largely functional." For SharePoint, "Accessing files stored directly in SharePoint Online remains functional."
- Downdetector logged more than 129,000 Outlook reports by 1:09 PM per Variety, roughly 130,000 at peak per CNBC, Today and Newsweek.
Timeline, from Microsoft's posts
- 11:20 AM First admin center entry. Investigating. Microsoft's first X post followed at 12:54 PM, per Variety.
- 12:45 PM "connectivity errors occurring on a portion of Exchange Online infrastructure which receives and processes users' access requests."
- 1:19 PM Affected infrastructure "is experiencing elevated load across multiple regions." Restarting and repairing it. This one reached X at 1:35 PM, per Variety.
- 1:29 PM Some US traffic redirected off the affected infrastructure. Other services split out to MO1550821 (BleepingComputer).
- 1:43 PM Servers are "becoming intermittently unavailable, which places additional strain on the remaining infrastructure." Outlook traffic gets spread across worldwide infrastructure (Variety quotes the X version). At this point they are still looking for "the source of the increased traffic" (Deadline quotes the earlier "unexpected increase in a specific type of service traffic" line).
- 2:06 PM "we've paused all deployments and configuration changes while we review recent changes for any correlation to this event."
- 2:18 PM The important one: "the elevated load isn't driven by an overall increase in traffic, but by a large number of servers being unexpectedly unavailable, which is shifting load onto the remaining infrastructure."
- 2:43 PM Servers are "failing automated health checks, which is causing them to become unavailable, and this may be related to a recent change to a component involved in processing those checks."
- 2:57 PM Walked back 14 minutes later: "we haven't identified any recent changes that correlate with the start of impact." The failed checks "may originate from a specific component on the affected servers that's responsible for processing them." They bypass that component on a subset of servers, stop more servers from being pulled out of service, and start looking at "elevated processor usage."
- 3:38 PM "many of the affected servers have become unresponsive, which is preventing us from confirming whether the change is effective." Restarting them with the change in place.
- 3:56 PM First progress: "an alternative health check method" let a subset of servers "successfully return to service." Rolling it out more broadly, plus an automated process to keep restarting servers that hang. And the change question is back: "a small number of configuration updates that we're investigating as potential contributors."
- 4:31 PM Still restarting and validating servers. "assessing whether reverting the small number of recent configuration updates we identified could help accelerate recovery."
- 5:00 PM Setback: "our actions taken thus far have not provided the anticipated relief." Also: "a core process on the affected servers is becoming overwhelmed, which causes the servers to stop responding and contributes to the capacity loss." New plan is to fully isolate a portion of the infrastructure, restart it, and gradually reintroduce traffic before doing that more broadly. They are also "evaluating whether isolating traffic within each region may allow affected regions to recover independently."
- 5:57 PM First good news: "We've begun returning a portion of the affected servers to service, and they're remaining stable so far." Microsoft reports "limited recovery in some regions" and is adding "protective traffic management measures in a specific location," to be expanded if they help. Expect "a gradual recovery as their clients reconnect to the service."
- 6:59 PM The recovery is not holding yet: "While the servers we've returned to service initially remain healthy, we've observed that they can become unresponsive again once they begin receiving traffic." Microsoft is "temporarily limiting which types of traffic can reach the affected infrastructure in specific locations" to see "whether a particular type of request is contributing to the impact," and is "rebuilding a subset of affected servers using a previous software configuration."
- 8:06 PM Microsoft names the choke point: "a process on the affected servers responsible for establishing secure connections becomes overwhelmed, causing the servers to stop processing requests." A configuration change to fix that is being validated on a subset of servers before a wider rollout with restarts. And for the first time it ties a change to the start: "we've begun reverting a recent network configuration change that correlates with the start of impact." That reverses the 2:57 PM line about finding no correlating change. The traffic limits from 6:59 PM are "now in effect in specific locations," with servers there handling "a narrower set of connections" before more traffic is let back in.
- 8:45 PM "We're seeing positive results with our efforts to restart affected infrastructure." Microsoft is monitoring those servers for stability. "Remediation actions will be expanded as we confirm stability with the changes applied."
- 9:09 PM "We're applying infrastructure capacity improvements where possible in an effort to help support traffic in regions coming online." Microsoft is still watching the results of the earlier changes before deciding what to roll out across all affected regions.
- 9:54 PM Setback. Microsoft says it has "confirmed that change is not taking as expected" and is "identifying alternate workstreams that can be applied to increase capacity in the face of sustained traffic management imbalances."
- 10:34 PM Microsoft is reviewing traffic flows and affected infrastructure "to identify opportunities to optimize routing" and is "verifying the reasons for service failures so we can develop specific actions to improve stability." It also says "there is no indication this is the result of malicious activity."
- 11:10 PM Microsoft is "adjusting traffic protocols, redistributing traffic, and applying strategic traffic blocking measures." It reports "partial recovery across regions in which we've applied these adjustments," but says "these actions aren't resulting in sustained recovery" and is reviewing what else it can do.
- 11:59 PM Microsoft says that as it applied its mitigations it saw "a surge in outbound call volume from our routing services to our backend servers," and is "continuing to observe heightened volume due to traffic rebalancing actions that we have taken throughout this event." It is working on "multiple workstreams to pinpoint the underlying cause of this traffic influx." Actions: "scaling up virtual machines to increase capacity and reduce load," and trying to "force sync for our recent network change" on machines that are "stuck in a bad state, preventing our changes from syncing as expected."
- 12:43 AM Still not recovered. Microsoft is "taking a measured approach to identify and repair affected infrastructure" and continues to "pursue multiple workstreams to pinpoint the underlying cause of the traffic influx and advance the recoverability of the backend servers." It says "we are not aware of any malicious activity that is causing this incident," and "Once mitigated, we will provide full details of the root cause, defense-in-depth repair items, and more within a Post-Incident Report."
- 1:16 AM Microsoft is "evaluating traffic controls that can be applied to increase capacity and stability," including "traffic throttling, protocol adjustments, and failover routing." It is also "expanding internal testing to track activity within a contained environment." It says again it is "not aware of any malicious activity that is causing this incident."
- 1:56 AM Microsoft is confirming "that developed patches have been applied and working as intended after infrastructure reboots are completed," and says "we have observed marginal success." It is also "taking action to expand capacity through the utilization of Virtual Machines."
- 2:25 AM Microsoft reports "positive results for server stability after applying an upstream authentication/routing configuration component patch to a portion of our infrastructure," and is "working to expand it to servers supporting active regions coming online."
- 3:06 AM Microsoft is "continuing to see increased server capacity stability after applying the upstream authentication/routing configuration component patch" and has "initiated its expansion systematically to focus on infrastructure supporting regions currently within or entering peak business hours."
- 4:55 AM Microsoft sees "early signs of stabilization, but service capacity has not yet returned to normal levels and health remains variable as traffic is carefully reintroduced." Teams are "cycling affected infrastructure through traffic drain, restart, remediation, and staged return-to-service." It is "extending the email timeout window to 48 hours to help protect delayed messages."
- 6:01 AM Microsoft reports "recovery actions completed across portions of the affected infrastructure in the Asia Pacific region" and says "Traffic is being gradually restored," but "some users may continue to experience intermittent issues." Next focus is "extending these recovery actions across affected infrastructure in other regions."
- 7:00 AM Microsoft reports "encouraging recovery progress in Europe, with service availability and email delivery performance improving across a growing portion of affected Microsoft 365 infrastructure." "Japan remains comparatively more stable, while North America is also improving; however, some users may still experience intermittent issues as recovery actions continue." The "extended 48-hour email timeout window remains in place." It says "we have no evidence of malicious activity associated with this incident" and will share "a full Post-Incident Report with root cause and defense-in-depth actions after mitigation is complete."
- 8:00 AM Microsoft reports "continued recovery progress across affected regions, with broader improvements now observed in North America and sustained improvements across much of Europe," and "encouraging recovery signals in Asia Pacific." It warns "some users may continue to experience intermittent access issues or delayed email delivery as mitigation progresses across the remaining affected infrastructure." It has added that "on-prem mail flow may be affected if it goes to or from Exchange Online." It is working "toward full mitigation" and has not declared the incident resolved.
- 9:01 AM Microsoft reports "improvements in service availability, with email queues significantly reduced and approaching normal levels." It adds "Some users may still experience intermittent access issues or delayed email delivery as recovery continues." It has "implemented additional mitigation measures" and is "restoring the remaining affected infrastructure while investigating residual service issues." It has not declared the incident resolved.
- 9:57 AM Microsoft says it is "observing sustained improvements in service availability as recovery progresses across affected Microsoft 365 services," but "some users may still experience intermittent connectivity issues or delayed email delivery." Its efforts are "increasingly focused on resolving the remaining issues and restoring consistent service for all affected users," and it is "validating recovery across different connection methods." It has not declared the incident resolved.
- 10:53 AM Microsoft says it remains "focused on addressing residual impact and ensuring reliable service restoration," and continues to "assess recovery across various connection methods and closely monitor service health to verify sustained stability." It has not declared the incident resolved.
- 12:57 PM Microsoft says "We are seeing broad recovery and stability across our services." It states "This was a technical issue not related to cyber or AI / agents," and that "The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change. We have addressed the underlying issue and continue to restore service to remaining customers and regions." It adds "Given the scale of demand, we restored services gradually, validating stability at each stage," and "We will provide a full post-incident report once our review is complete." Both incidents are still listed as degraded and the formal Root Cause field still reads N/A.
- 3:02 PM Microsoft says "As infrastructure stabilizes, we're still draining queued messages and rebalancing traffic, so customers may see improved sign-in/access while delivery and processing continue to catch up." It adds "We do not have an ETA for full recovery yet. Service access can recover before mail flow fully normalizes." Both incidents are still listed as degraded.
- Fri 5:00 PM Microsoft says "We continue to observe sustained service recovery across the previously affected environments." It is "monitoring for an extended period to ensure no additional mitigative actions are required" and is "monitoring the backlog of email messages accrued during this event to ensure all process successfully before considering the issue resolved." Both incidents are still listed as degraded.
- 9:13 PM Microsoft says "We've confirmed continued recovery across previously affected environments." It will "maintain monitoring to ensure the mitigation actions taken remain effective and track processing of the email backlog accumulated during the event." It adds "We believe the majority of these scenarios have been addressed, with work continuing on any remaining scenarios." Both incidents are still listed as degraded.
- Fri 11:15 PM Microsoft posted its final status for EX1550543. It says "This was a technical issue not related to cyber or AI / agents." and "The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change." It lists the incident as starting Thursday, November 12 at 10:05 AM UTC and ending Friday, November 13 at 4:10 PM UTC, and labels that cause a "Preliminary root cause." It adds "We'll provide a preliminary Post-Incident Report within two business days and a final Post-Incident Report within five business days." Both EX1550543 and MO1550821 now show Service restored. Nov 16: Microsoft has since revised this entry: the start now reads 4:20 PM UTC and the preliminary root cause has new wording. See the update at the top.
Both incidents now show Service restored. Microsoft's final entry gives a "Preliminary root cause" and lists the incident as running from 4:20 PM UTC on Nov 12 (11:20 AM ET; first listed as 10:05 AM UTC, revised by Microsoft, see the Nov 16 update) to 4:10 PM UTC on Nov 13 (11:10 AM ET). Microsoft published a post-incident report to admin centers on Nov 15; its text isn't public.
My read (mine, not Microsoft's)
Update, Nov 13, 11:45 PM ET: Microsoft has closed both incidents and given its cause in its own words (Microsoft first labeled this its "Preliminary root cause"; see the Nov 16 update for the current wording): "The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change." It also says "This was a technical issue not related to cyber or AI / agents." That is Microsoft's preliminary account, not a final one. It has promised a preliminary Post-Incident Report within two business days and a final one within five, and the detail (which upgrade, which change, why it took this long to recover) waits on those.
What I wrote while it was happening, left as it was: servers started failing automated health checks and got pulled from rotation. The rest took the extra load, and a lot of them went unresponsive too. That is a capacity cascade. At 8:06 PM on Nov 12 Microsoft said it was reverting "a recent network configuration change that correlates with the start of impact," after suspecting a change at 2:43, not finding one at 2:57, and putting it back on the table by 3:56. Microsoft's preliminary root cause also points at a network change, so that lead looks like the right neighborhood. The Post-Incident Report should say whether it is the same change.
Nov 16: Microsoft's preliminary root cause fits what was visible from the outside. On Nov 12 the Outlook front ends answered ping but most HTTPS connections to them timed out, and at 8:06 PM Microsoft said "a process on the affected servers responsible for establishing secure connections becomes overwhelmed." Servers that "could no longer accept new connections" is that same picture. The health-check failures I described as a capacity cascade are not mentioned in the preliminary root cause, which neither confirms nor rules that part out. Which maintenance, what the latent bug was, and why recovery took about a day are questions for the report itself.
What it does and doesn't look like
- Not a traffic flood, per Microsoft's own 2:18 PM update. Microsoft addressed the attack question directly in its 10:34 PM update: "there is no indication this is the result of malicious activity." Before that it had simply never attributed this to an attack, and none of the coverage I've read (ABC, Global News, SecurityWeek) reports one. Microsoft's statement to CNBC only says it is "investigating reports of users experiencing issues with Microsoft 365 services, primarily issues with Exchange Online." CISA's alerts page had nothing on it when I checked at 6 PM ET; the newest alert was dated Nov 10. On Nov 13 at 12:57 PM Microsoft went further: "This was a technical issue not related to cyber or AI / agents." It repeated that sentence in its final status when it closed the incident at 11:15 PM.
- Not Google. The Workspace status dashboard shows no incidents today, so Gmail showing up on Downdetector is most likely people reporting "my email is down." Cloudflare's status page shows nothing broad either, and AWS says it is "operating normally" and that outage reports about it are inaccurate, per CNBC.
- Not DNS, from where I sat on Nov 12. The Outlook hostnames resolve to normal Microsoft front-end IPs on 1.1.1.1, 8.8.8.8 and my ISP's resolver.
- Not routing, from where I sat on Nov 12. The front-end IPs answer ping with no loss, but most TCP 443 connections to them time out. Packets arrive; the servers aren't taking work. It has flapped all afternoon between zero and three of four test IPs accepting connections.
- Sign-in looked fine from here on Nov 12. login.microsoftonline.com answers normally. Graph answers too, but mailbox calls come back 502/503/504. One network, my own tests, so weigh it accordingly. Microsoft's 7:00 PM scope text lines up with it: Graph "API requests that access Exchange Online data, such as mail, calendar, and contacts, may time out."
If you're fielding the tickets
- Don't have users reset passwords, rebuild profiles, or reinstall Office. Nothing client-side fixes this.
- Inbound mail: Microsoft's 7:00 PM scope text for MO1550821 now says "email delivery may be delayed." It has not said anything about mail being rejected or lost. I have not tested it end to end. If senders are getting temporary (4xx) errors, their servers retry on their own schedule, typically for a day or more, so late is more likely than lost. Check message trace and your own queues once you can get in.
- Microsoft closed both incidents at 11:15 PM ET on Nov 13 and no further status updates are scheduled. Its final status still says it continues "to restore service to remaining customers and regions," so a few stragglers are possible: delayed mail still arriving and some users reconnecting later than others. If something is still broken on Monday, open a ticket and reference EX1550543. Watch the admin center for the Post-Incident Report.
Microsoft marked this resolved at 11:15 PM ET on November 13, 2026. Last checked against Microsoft's feed at 11:35 PM ET that night. Microsoft published a post-incident report to admin centers on Nov 15 (see the update at the top). I'll add a note if the report's text or a final version becomes public.
Edit, Nov 12, 7:45 PM ET: repointed several source links so each one opens the page that carries the quoted text, and noted where the X posts ran later than the admin center entries. No facts or quotes changed.
Edit, Nov 12, 8:05 PM ET: added Microsoft's 7:00 PM rewrite of the MO1550821 scope. The mail flow line used to say Microsoft had said nothing about delivery; Microsoft now says delivery may be delayed.
Edit, Nov 13, 1:10 PM ET: added Microsoft's Nov 13 cause statement to the "My read" and attack sections, marked my Nov 12 network tests as past tense, and replaced the overnight advice now that Microsoft reports broad recovery.
Edit, Nov 13, 11:45 PM ET: Microsoft marked both incidents Service restored. Correction: this post gave the EX1550543 start time as 11:17 AM ET on Nov 12, which is what Microsoft listed at the time. Microsoft's final status puts the official start at 10:05 AM UTC (5:05 AM ET) on Nov 12 and the end at 4:10 PM UTC (11:10 AM ET) on Nov 13. I also updated the "My read" and ticket sections to reflect the closure and Microsoft's preliminary root cause.
Edit, Nov 16, 1:57 PM ET: added Microsoft's post-incident report status and its revised preliminary root cause, and corrected the official start time from 10:05 AM UTC to 4:20 PM UTC after Microsoft revised its final status. Earlier wording is left in place and marked.
Sources
- Microsoft Service Health text for EX1550543 and MO1550821: UPenn's public mirror (times in UTC) and TechRadar's live blog
- @MSFT365Status on X, status.cloud.microsoft, CISA alerts
- Press: BleepingComputer, CNBC, Variety, Forbes, Deadline, ABC News, Global News, SecurityWeek, Today, Newsweek
- DNS, ping, TCP and Graph results are my own tests from one network