Is AWS down? Live updates as Amazon disputes outage and users disagree
AWS is disputing the outage, but users disagree
Reminder that everything is fine, officially
Reminder that, despite the now-dwindling Down Detector reports, AWS is still considering all systems operational, with the last issues reported yesterday afternoon.
Be sure to double-check whether any issues you're having aren't actually being caused by Azure and the ways it's connected to AWS.
Azure and AWS are connected
Azure and AWS are interlinked in some key ways, so the Azure outage could be the cause of the AWS outage reports on Down Detector.
If Amazon is correct and AWS isn't down, then this could be an interdependence thing — many companies use both AWS and Azure servers for different services.
Either way, for people who rely on Azure- or AWS-backed services, it probably doesn't matter which one is causing the problem — all that matters is they can't do what they need to.
Down Detector comments...
It's not just emails from users telling us that something is wrong. The Down Detector comments section is filled with users complaining that services they use that rely on AWS are not working.
"CalCareers is down now. This is exactly what happens when one company monopolizes everything! When AWS fails, everything else falls like dominoes, and the entire system is shut down," one user wrote.
Another said, "I still can't access my advance payroll service, that's is using AWS, so for me it's still down. The error states due to a technology issue with a partner."
On the other hand, one Down Detector user said, "AWS is fine. I think folks are mixing up AWS and Microsoft Azure. Not the same. Azure is down," which is a theory I put forth earlier. Another responded saying it was down but is fixed now.
There are a lot of different perspectives on this one. Whether it's Azure confusion or an actual AWS outage remains to be seen.
Whether AWS is down or not is debatable but..
We're still getting emails from readers regarding issues with AWS-connected services. For example, one user said, "Amazon definitely had issues. Fire TV couldn't connect to anything for over and hour. Just kept saying unable to connect to servers."
AWS powers the backend of Fire TV, so there definitely could be a connection there.
Another user said, "We just tried to call U-Haul in Seattle. They said their systems were down and they can't make reservations." Based on my research, U-Haul uses AWS for its systems, further making us wonder what's actually happening here.
Amazon proper received some reports, too
We also saw Amazon spike on Down Detector, though it was much less significant that the spike we saw on AWS. I've looked at Amazon several times while this alleged outage was happening and it loaded for me without issue.
Can this many people be wrong?
The number of reports on Down Detector is declining, but many people are still experiencing problems. A 6,000-person peak with 2,500 as of this writing is substantial, so for nothing to be wrong would be surprising.
Could people be mistaking the Azure issues for AWS and reporting on the wrong thing? While it seems unlikely that that many people would make such a huge mistake, it is the internet, and you never know for sure.
Obviously, the AWS status page is the most reliable place to go, as Amazon said in a statement to us, but it's interesting to see how many reports came in.
A developer chimes in
Davi is a "DevOps, Infra, or Platform engineer for a startup," and he's emailed in saying "over the last two hours, we've been experiencing instabilities with their SPOT instance capabilities, and on the last hour, a total interruption. The following photo is from one of our EKS clusters." Specifically, he's talking about the US-EAST-1 region.
To break this down, think of the AWS region as a multi-level parking garage, and the "UnfulfillableCapacity" error being like an attendant telling you that "there isn't a space matching your exact needs."
Davi has also shared the following from a site that uses AWS' official API to map global spot instance errors. To verify for yourself, go to this site and change the selector to AWS.
More people are mailing in
Barry from Torrance, CA has mailed in to say that "Starbucks and Einstein Bagel both have non operating apps. Further a Starbucks employee told me that in the store he was not able to process orfers earlier today."
Starbucks uses AWS to run its customer rewards and mobile commerce platforms. Einstein Bros. is a little more tricky to find concrete information on this, but from what I've been able to glean from job sites and whatnot, it looks like a mix between AWS and Azure services.
So why the spikes?
It's an interesting question, because Down Detector is a user platform for reporting outages, and this is the first time in my time at Tom's Guide that they have been in direct opposition to the service status of a system.
BREAKING: AWS disputes the outage
AWS has just reached out to us with a statement: “AWS is operating normally and this reporting is incorrect. The only resource on the internet that provides accurate data on the availability of our services is the AWS Health Dashboard.”
True enough, according to the AWS Health Dashboard there are currently no problems at AWS. Which makes this huge spike on DownDetector even more unusual. Rest assured, we're looking into what's going on right now.
Microsoft services are down too
You may have noticed on Down detector that it's a lot more than just AWS services, Microsoft systems seem to be down too. For something like Outlook, I get it — some companies run their email servers through AWS, but to see Minecraft, Halo Reach and Copilot go down too? This is an Azure problem.
U.K. Check
Just like last time, a lot of the UK infrastructure is seeing user reports increase too.
