Another M365 Outage?

Posted by _-RustyShackleford@reddit | sysadmin | View on Reddit | 46 comments

I'm not seeing any outages on my end and so far I haven't heard from any users (it IS 7am, so that's not a shock), but is anyone seeing impacts from this alert?

Users:

Users are unable to access Microsoft 365 apps.

Scope of Impact:
Impact is specific to some users who are served through the affected infrastructure, attempting to access Microsoft 365 apps.

Updates
We're continuing to review service monitoring telemetry to isolate the source of the issue and establish a fix.

Oct 13, 2025, 6:18 AM EDT Next update by:

[-]

That-Acanthisitta572@reddit

I know all the copilot shit but I didn't think MS would start vibe coding the Admin centres THIS quickly... Because all these issues and outages really do make me feel like it's just some AI pushing shit at this point.

[-]

Helpjuice@reddit

Not privy to what is going on internally but outages do and will happen even at the AWS, Azure, GCP, Oracle so there should be contingency plans in place to keep operations running in case of outages. Do what the government did and run out of multiple cloud providers.

[-]

yankmywire@reddit

Or repatriate those workloads to on-prem. That way.. you're the one in control (and likely the cause/effect) of the outage.

[-]

Fallingdamage@reddit

If I ran Exchange on prem and had as many outages as microsoft has with Azure, I would be on unemployment right now.

[-]

Helpjuice@reddit

True, and it is not locked in by only being available from one cloud provider. You want to increase coverage and scaleability you spin up more exchange services in other data centers. Though, there are other options that do not involve using Microsoft that scale and allow you flexibility in how much availability you have based on what the company invests into making it happen which can cost less than using cloud solutions.

[-]

RCTID1975@reddit

it is not locked in by only being available from one cloud provider.

Sure, but it just shifts the risk. Now you have increased security risks, increased infrastructure risks, ISP risks, Internet backbone risks, building risks, personnel risks, talent/skill gaps and risks.....

[-]

wideace99@reddit

For Micro$oft there is more money to squeeze from cloud compared to on-prem :)

The "wise" business decision to go on the vendor lock-in way based on cheaper costs + official support seems not only to make the budge bleed + a crappy support but also have uptime problems.

[-]

Frothyleet@reddit

I don't know if there are any reasonable ways to multi-cloud most of the M365 stack. Azure, sure, you can replicate its functions and services in AWS or GCP. And for DR purposes, you should have a fallback plan for M365.

But you're not going to be able to just stand up Exchange Online - AWS Edition and keep your workflows and users happy.

[-]

wideace99@reddit

Maybe you should take a look in the FOSS world, where the biggest email servers are not hosted by Exchange and still have redundancy & scalability :)

[-]

Frothyleet@reddit

I'm well aware that Exchange is not the only MTA out there, but I struggle to see how you can reasonably build redundancy with your M365 stack in other providers.

Again, DR? Sure. Restore your M365 backups to Google or your homebrew MTAs.

But nothing you'd just casually flip the switch on during downtime.

[-]

wideace99@reddit

My initial post is about to completely replace your vendor lock-in stack with FOSS.

But nothing you'd just casually flip the switch on during downtime.

It's possible to have one backup MTA (or multiple for redundancy & scalability) and an automatic switch (script) to be able to receive/send email while your Micro$oft cloud is not functional, but you won't be able to access already stored email's since it's stored in your Micro$oft cloud which is down.

Having a complete FOSS solution redundant & scalable has no dependence on Micro$oft up/down status.

[-]

Frothyleet@reddit

Right, so while I appreciate your enthusiasm and antagonism towards MS, there's not a realistic way for the vast proportion of orgs who are invested in M365 (or even Google Workspace) to build anything close to seamless tolerance for those services being down.

The best practices solution is simply what everyone should already be doing, which is SaaS backups outside of M365 (whether that's Veeam, Dropsuite, or a gazillion other options).

[-]

wideace99@reddit

In this case, just continue to enjoy Micro$oft services :)

[-]

Frothyleet@reddit

I'm not sure enjoy is the correct verb, but I'll be doing somethin' with them

[-]

RCTID1975@reddit

the biggest email servers

Right. 99.9% of the people in the world aren't running those, nor have the budget for that.

CAN it be done? Sure. Is it feasible or logical? Not typically

[-]

wideace99@reddit

This is not for "people" but for companies who already are paying big money to Micro$oft.

Not happy about the change ?

It's no problem for me, my servers (not only MTA) are working fine.

Continue to enjoy micro$oft services :)

[-]

Silent_Rule_S@reddit

And Intune? And Sharepoint? And Teams? And Entra/AAD?

[-]

wideace99@reddit

There are equivalents in the FOSS also for them :)

Of course, they are under different names and also have variously plus/minus functionality and the best part is that only experienced Linux CLI sysadmins can glue them together :)

[-]

Silent_Rule_S@reddit

ah yes we bow to your superior knowledge :)

Tell me, how may enterprises in the real world actually run a stack like that? :)

[-]

wideace99@reddit

Just the little ones like: Google, Yahoo and many other commercial mail services :)

[-]

Helpjuice@reddit

The key here is not locking yourself into a single provider for your most critical business operations.

[-]

Frothyleet@reddit

Sure, that'd be ideal. I'm interested in how your org has done that for productivity and collaboration stack.

[-]

Silent_Rule_S@reddit

So, run two parallel office, collaboration suites and authentication/login and MDM platforms?

Does ANYONE do that?

Don't think you understand what O365 is.

[-]

Helpjuice@reddit

If you choose to go with one vendor you accept the risks of outages with just that vendor. If you want to increase your availability options you cannot lock yourself into one provider of options - M365.

Example would be other organizations not using M365 and hosting their collaboration suite using other products and services that they maintain internally.

[-]

Silent_Rule_S@reddit

So an absolute mess of on-prem email, open source wikis, LDAP/Keycloak/Samba and Libreoffice?

Like I love open source, Firefox especially as I can block ads but dawg 💀

[-]

wideace99@reddit

Only those who are paying right wages for professionals :)

The rest can "enjoy" cheap SAAS + it's "wonderful" tech support (aka Rajesh) + outsource MSP (aka another Rajesh) + internal support (employees who accept the cheapest local wages) :)

[-]

Silent_Rule_S@reddit

I also enjoy being racist and h8 Indians in tech. Respect.

[-]

Helpjuice@reddit

Many just use open source products and Microsoft's own products. Some do not want their data in the cloud, or only want a subset of it there. The whole there is only M365 doesn't make any sense as this is exactly how the world did hosting before the managed service was available from Microsoft and how many organizations still do it today using Microsoft or other vendor products.

[-]

RCTID1975@reddit

This is why we have an entirely separate office building with a second set of employees. We keep them on a different planet in case the earth implodes one day.

Your proposal here is nonsensical

[-]

EnragedMoose@reddit

Azure is notably shit though.

[-]

ForeverOk5504@reddit

Microsoft is crazy today

[-]

Fallingdamage@reddit

Maybe copilot was given too much freedom on the backend.

[-]

Sufficient-Class-321@reddit

Yeah, I clocked this earlier. I've also noticed that Admin Center has been crashing out and giving me all kinds of hell today too

[-]

Fallingdamage@reddit

Yeah. Lots of random errors and seemingly petty things. For instance, I can add users to some distribution groups, and other distribution groups refuse to accept any additional members for some reason.

[-]

lart2150@reddit

in chicago https://admin.microsoft.com/ is spinning but OWA and entra are fine.

[-]

_-RustyShackleford@reddit (OP)

Honestly, right now I'm not sure what is an Issue and what is due to my Security team trying to switch/rotate admin accounts.

[-]

Ambitious-Airport360@reddit

Same with Admin Centre - glad it's not just me.

So miuch for "we've fixed the issue" - Microshit at it's finest.

[-]

bjc1960@reddit

We are seeing DNS issues with Entra Private Access. IP address works.

[-]

christurnbull@reddit

Autopilot was being strange again today

[-]

Caleth@reddit

Ime autopilot is often strange refusing to load some program or hanging on a profile setting.

[-]

RCTID1975@reddit

refusing to load some program

That's not autopilot. That would be intune.

[-]

RCTID1975@reddit

served through the affected infrastructure

This is why location matters in outage posts, and why these arguments of "There's an outage multiple times a week!!!!" are ridiculous.

[-]

TravellingLuchador@reddit

Had an issue this morning where a user was requested to sign in using a domain we haven't seen before. This domain has no DNS records(MX, SPF, etc) and can't find anything in our system.

Anyone else have an experience like this?

[-]