00:01All right, 17 hours. In fact, more than 17 hours and counting.
00:05That's how long Microsoft's engineers have been at it.
00:09And by the company's own admission, they have had marginal success.
00:13Now, you already know the story because if your outlook was stuck spinning last night, it wasn't your Wi-Fi.
00:20It was the entire Microsoft network from Copilot to MS Teams where a lot of meetings take place to even
00:30Outlook,
00:31which is one of the places where almost all companies operate when it comes to their emailing system as well.
00:39Now, how did it start? It started in the U.S. on Thursday afternoon.
00:43For us, it was around 9.30 last night.
00:45Outlook went first and then Teams started acting up and then it was Copilot.
00:51Within about an hour, people in 74 countries were reporting problems.
00:56In India, the complaints started coming in around 11.30.
00:59So, by the time most of us were logging off for the night, Microsoft already had a global problem on
01:05its hands.
01:06Now, how big is this? Honestly, nobody can really tell you yet, not even Microsoft.
01:11It hasn't said how many users were hit.
01:13What we have is the down detector and the down detector only counts people who were bothered, you know, who
01:21even bothered to complain over there.
01:23So, in U.S. alone, people who complained were about 1.3 lakh.
01:27These are only people who have filed a complaint with Microsoft that their server has been hit and their Microsoft
01:33Outlook is not working.
01:34You will hear another number online which is doing the round, which is 430 million.
01:39That's how many people use Microsoft 365 apps worldwide.
01:44430 million people use Microsoft 365 apps worldwide.
01:49And it is not how many are really affected.
01:52But, you know, it just tells you the scale of outage that people are really seeing and disruption that is
01:58really happening.
01:59So, what actually broke?
02:00Now, Microsoft says that some of its servers suddenly went down.
02:03The ones still running had to take everyone else's load and they, of course, buckled.
02:08That's why.
02:09And here's why this has sort of dragged on.
02:12When engineers brought servers back, some crashed again as soon as the traffic hit them.
02:18So, the big question at the moment that everybody is asking is, because we do not have a clear answer
02:23for Microsoft is, is this a hack?
02:25Microsoft says no.
02:28Charles Lamana, one of the senior executives says that there's no sign of any cyber attack.
02:33The real root cause is still under investigation.
02:36So, while Microsoft is ruling out any kind of a hack, it is not giving us a reason on what
02:41really happened.
02:42So, the question then up next that comes to mind is, why should this bother you beyond the missed meeting?
02:47Some might actually enjoy it.
02:48Well, because it wasn't just inboxes.
02:51Banks, government offices, even courts in several countries reported disruption.
02:57And co-pilot, the AI that's meant to sort out your email for you couldn't reach it either.
03:03Your AI is only as smart remember as your server.
03:06It runs on.
03:07So, everything is in a bit of a limbo at the moment.
03:11But this is not the first time.
03:12We did some research to find out how long have outages really lasted.
03:17And this is what the figures tell us.
03:20The last one that we heard of was actually, in fact, let's go back to 2021 and start there.
03:24In October of 2021, Facebook, WhatsApp and Instagram went dark for about six hours.
03:30In July of 2024, one had CrowdStrike update knocked out just about 8.5 million Windows machines completely.
03:39Flights were grounded.
03:40Hospitals and banks were hit.
03:42And some organizations even, you know, took close to 10 days to fully recover from that outage.
03:48And we learn the same lesson every time, right?
03:51And every time we go back to putting everything in one basket.
03:53Microsoft says it will publish a full root cause once this has been fixed.
03:59It will be reading, you know, all of that very closely, of course.
04:03Because the bigger question isn't about when Microsoft really gets back or Outlook gets back for all of us.
04:10It is this.
04:11When a company that runs half of the world's offices goes down for a full day, who pays for that
04:20lost day?
04:22Right now, it's just the users.
04:26We'll have to wait and see how that story really progresses.
04:37In fact, we have an update coming in now.
04:38Microsoft has given an update.
04:40It says on its server outage.
04:42It says we have made additional progress towards service recovery with recovery actions completing across portions of the affected infrastructure
04:51in the Asia-Pacific region.
04:54Traffic is being gradually restored and we are seeing improvement in key customer scenarios as well.
04:59Focus is now on extending recovery actions across effective infrastructure in other regions as well.
05:06So Microsoft there, since the sheer volume of numbers comes from the Asia-Pacific region as well, they've made it
05:12a point to underline that,
05:14to say that they have made significant recovery, additional progress to be precise here.
05:20Additional progress has been made during the service recovery, specifically in the Asia-Pacific region.
05:26And once that is gradually restored, they will be looking at improving the scenario in other parts of the world
05:32as well.
05:36Alright, Saurabh Singh from Tech Today joins us for more details on these stories.
05:41Saurabh, there are three kinds of theories online.
05:43One, it's a hack.
05:46Number two, well, too much of AI and very little of engineers.
05:52This is why engineers, when you fire them, this is what happens.
05:56That's the second theory online.
05:57And the third theory really is that the servers simply do not have the capacity and they were not given
06:04additional support.
06:05The constant, you know, mechanism to ensure that their servicing happens has not happened
06:11because the speed of which AI is really growing.
06:14Are any of them true?
06:16Are none of them true?
06:17And how long are we expecting this to go on?
06:21So it's a very evolving thing.
06:23Microsoft itself hasn't given out the exact details.
06:26All of these three theories that you mentioned, they are, of course, actively being discussed online.
06:31But we do not have clarity as to, I mean, which one is true.
06:34Of course, Microsoft says that servers have been sort of overwhelmed and this is what causing the exchange online errors
06:43because of which Outlook and services that, you know, are used in communication, right, from calendar to scheduling, they are
06:51not working.
06:51It's been over 17 hours and the company is fixing.
06:56They made some improvements.
06:57They have informed that recovery is happening.
07:01It's going to take some time though, it seems.
07:04Some time.
07:06So they can only thank the calendar that it's the weekend.
07:09So they'll have a bit of a relief to actually restore this.
07:12But any idea then on is it really, when they say server capacity is limited, what exactly does that mean?
07:22Does that mean some servers were not repaired in time?
07:25Does that mean there was suddenly too much of a load?
07:28Which the second one seems unlikely because there's been no activity really that will get people online.
07:32It's not like chat GPT under some, you know, meme has crashed.
07:37It's not like that in this scenario.
07:39So what exactly does this mean? Server crashing?
07:42What could the possible reason be?
07:43So Microsoft has sort of given some clarity that that was not really the case.
07:49A second reason that they've also quoted alongside is that they pushed out a configuration update
07:55and which is when these errors started to happen and servers sort of, you know, started getting overwhelmed, so to
08:03say.
08:03They have not mentioned that, you know, their server's capacity is sort of compromised.
08:09But at the same time, the errors because of the configuration update that was pushed out could be a reason
08:14why the overwhelming happened, so to say, technically.
08:19All right, a bad update yet again.
08:21That's getting company after company and Microsoft falling prey to that.
08:25It looks like at least at the moment, you know, that's the case.
08:29Thank you so much for joining us.
08:30Let's also take a look at a report filed by Tech Today editor Cyrus to take you through the details
08:35of what really, you know, happened here.
08:40Microsoft is facing an outage globally and India has also been affected.
08:45Now, I'm talking about services like Microsoft 365, its AI co-pilot and most importantly, Microsoft Outlook.
08:52And it's causing businesses and services disruption in their operations.
08:56So what really has caused this outage?
08:59What is Microsoft doing to fix this issue?
09:01And when will users be able to use Outlook and other Microsoft services?
09:06Here is the full story.
09:11The trouble began on Thursday, November the 12th, when users started reporting problems with Outlook.
09:17Some could not log in while others couldn't send or receive emails.
09:21But the problem soon spread to other Microsoft services, including Teams, OneDrive, SharePoint, Co-pilot and Planner.
09:28That's because many of these services rely on Microsoft's cloud infrastructure.
09:33So when that ran into trouble, several connected services were affected as well.
09:42So what went wrong?
09:44Microsoft found that some of its servers had stopped responding.
09:47These servers handle millions of requests from users every day.
09:51While several servers going offline, the remaining ones had to handle more traffic than they could manage.
09:58To make matters worse, a system that helps establish secure connections also became overloaded.
10:05Think of it as a traffic jam caused by several roads suddenly closing.
10:09The remaining roads simply could not handle all the vehicles.
10:13Microsoft is also investigating a network setting change that happened around the time the outage began.
10:19The company says there is no indication of a cyber attack.
10:28Getting these servers back online has not been easy.
10:32Microsoft tried restarting them and shifting traffic to working servers.
10:36But some of the restored servers stopped responding again when users started accessing them.
10:43The company has since been working to bring them back gradually and prevent the same problem from happening again.
10:50In the latest update, Microsoft says a fix applied to part of its infrastructure is showing positive results.
10:57It is now testing this further and extending the fix to more servers.
11:05The disruption has been reported across 74 countries.
11:10According to Dundetector, reports of outlook problems peaked at over 1.38 lakh globally.
11:16And the impact goes beyond emails.
11:19Air India and Indigo reported problems with their email support services, although flight operations continued normally.
11:25Government services in Singapore and Philippines also faced disruptions.
11:31Businesses have struggled with communication, with one US company even delaying an earnings call.
11:36Even AI companies like OpenAI and Anthropic reported problems accessing some Microsoft 365 tools.
11:44At the time of recording this video, Microsoft has confirmed that it is seeing some positive results in terms of
11:50fixing the problems and releasing patches.
11:52However, there is no confirmed timeline on when all of these issues will be fixed.
11:57However, this is a sign of how the internet is so vulnerable and at the mercy of some of these
12:03big tech companies.
12:04This is Cyrus John for India Today.
