If You Can't Trust Amazon AWS......

[Paul Kelly ICHYB]

So, even Amazon can be hit!

Many big and small companies use it and their business will probably have been affected.

Ultimately, who can you trust, if anyone?

 

[fisicx]

As has been said many times before: all cloud means is someone else’s computer.

 

[Ozzy]

As has been said many times before: all cloud means is someone else’s computer.

Basically, yes, albiet in theory should be more easily scaleable

 

[fisicx]

Basically, yes, albiet in theory should be more easily scaleable

Indeed. But if someone cuts the bit of string connecting you to them all the scaling in the world isn’t going to help.

Can’t recall the exact number but there are a few key nodes that everything passes through. If your bit of this node falls over everything just stops.

 

[pentel]

Using amazon AWS is a get out of jail free for the IT department which is not the case if everything is on site.

 

[JEREMY HAWKE]

This should not happen and we cant get away from the cloud in 2025 and going beyond.

Outages ,system hacks and wifi / 4/5g drops out must not happen
The maintance and care should have the same focus as a passenger aircraft and they dont drop out the sky
We need to progress to the point where this is fool proof

 

[DontAsk]

If yo0u don't control the server you do not control the data.

 

[Paul Kelly ICHYB]

If yo0u don't control the server you do not control the data.

You can never totally control the server.

Even if you had a box in your house, you are still at the mercy of your internet provider and their resources!

 

[Ozzy]

You can never totally control the server.

Even if you had a box in your house, you are still at the mercy of your internet provider and their resources!

Plus, man-in-the-middle attacks, denial of service attacks, phishing attacks - all these nothing to do with the cloud and these plus more can affect on site infrastructure.

 

[DontAsk]

You can never totally control the server.

Even if you had a box in your house, you are still at the mercy of your internet provider and their resources!

If the server is on the box in your house then you can control it and who access it.

 

[WaveJumper]

And all this relies on that tingly little stuff that comes out a socket in your wall ............. currently over seen by someone called mad Ed

 

[Paul Kelly ICHYB]

If the server is on the box in your house then you can control it and who access it.

And if your internet connection goes down?

 

[DontAsk]

And if your internet connection goes down?

You the server with local access.

 

[fisicx]

So the AWS failure was DNS. It’s almost always DNS. A single point of failure for which there is no work around. It either works or it doesn’t.

 

[Chieh]

A capable architect builds systems that work across providers. Not by fearing downtime, but by learning to live with it.

 

[fisicx]

A capable architect builds systems that work across providers. Not by fearing downtime, but by learning to live with it.

Not sure that would apply in this case.

 

[johndon68]

A look at what happened.

- YouTube

Enjoy the videos and music that you love, upload original content and share it all with friends, family and the world on YouTube.

youtu.be

 

[DontAsk]

From the Register https://www.theregister.com/2025/10/20/aws_outage_amazon_brain_drain_corey_quinn/

TLDR: "At the end of 2023, Justin Garrison left AWS and roasted them on his way out the door. He stated that AWS had seen an increase in Large Scale Events (or LSEs), and predicted significant outages in 2024. " albeit a year later than predicted.

 

[DontAsk]

ROTFLMFAO https://arstechnica.com/gadgets/202...-sleepers-hot-and-bothered-during-aws-outage/
Some people really do have more money than sense.

 

[ThatDevAaron]

So, even Amazon can be hit!

Many big and small companies use it and their business will probably have been affected.

Ultimately, who can you trust, if anyone?

lol

AWS is not a protection provider, they are notorious for being absolutely terrible in this regard.

I always say, don't use AWS, Azure, Gcloud for your shit. They are only good for very very particular things.

But as they say; A jack of all trades is a master of none

 

[HFE Signs]

I guess one advantage with a larger host such as AWS is that so many people are affected it will mean you don't have to try to speak to someone and it will usually be quickly resolved.

 

[ThatDevAaron]

I guess one advantage with a larger host such as AWS is that so many people are affected it will mean you don't have to try to speak to someone and it will usually be quickly resolved.

Except; using the same infra as most businesses will mean, when people try to take AWS offline to affect the larger brands, you will be affected as a side-result.

Not to mention, AWS is notoriously shit with ddos prevention.

 

[fisicx]

I guess one advantage with a larger host such as AWS is that so many people are affected it will mean you don't have to try to speak to someone and it will usually be quickly resolved.

Maybe so but most UK small businesses don’t need to use AWS or any other cloudy hosts. There is very little benefit in doing so.

 

[fisicx]

An update here with a link to the Amazon incident analysis:

AWS outage: Myths vs reality

Column: AI wasn't the cause, and multi-cloud is for rubes

www.theregister.com

 

[Karimbo]

If the server is on the box in your house then you can control it and who access it.

you probably have less control than you think. you do not know how good your srcurit is. could easily misconfigure the server and leave access wide open for attacks.

 

[DontAsk]

You can’t trust Microsoft Azure either.

 

[Paul Kelly ICHYB]

Yup - no one is immune!

Microsoft Azure outage: Websites come back online

Microsoft 365 and its Azure cloud computing platform were hit with DNS issues, similar to the recent AWS outage.

www.bbc.co.uk

 

[DontAsk]

I keep saying it, or similar, but don't put all your eggs in a cloudy basket,.

 

[EnterprisePro]

Well, this could all happen with any cloud provider. As the saying goes, "don't keep all your eggs in one basket". We've setup our infra with a multi-cloud and hybrid-cloud approach. If one region in one provider has an outage, we're fine. From what I gathered in AWS, it was an issue affecting one region only, so even within AWS (and many other cloud providers) you can deploy your infrastructure in multiple regions to prevent service related issues to your own customers.