As a Cyber Security CEO said to me on Monday, "This may just be a genuine cock-up, but take out AWS and Microsoft Azure and half the economy stops...", UK Gov treating cyber resilience in the supply chain as part of National Security....if M&S and JLR are so vulnerable, God help the smaller companies.
Can't get to my money. . 😭😠
- Thread starter Sinistral
- Start date
я согласен
Half? The rest would go out in sympathy. Total collapse.
The whole reason for the "internet" was supposed to be resilience. We have probably never been more vulnerable.
TringSaint
Well-Known Member
It’s interesting as the core banking platform I used to sell, which sits on aws, was supposed to use the UK region and the two London AZ’s for dual node high availability computing, yet one such UK bank went down yesterday, so either it’s not just a US issue, or the non US EC2 regions are tied to the US ones very closely!!
There will be a few worried faces in the risk departments of a host of banks today, waiting to see if the PRA pays them a visit!!
There will be a few worried faces in the risk departments of a host of banks today, waiting to see if the PRA pays them a visit!!
I don't think that will be necessary - more likely it will be offerings similar to cloud at customer and sovereign cloud.
I can't speak for AWS, but it may depend on the exact flavour of cloud that was being employed, i.e. was it deployed on regular public cloud data centres, or dedicated/sovereign cloud?
If the former, then there's the chance that even though the data centers were in the UK, hence meeting the data residency requirements, the management of the data centres could still have taken place using services and resources located Stateside. These potentially could then have been affected by the outage.
If it was dedicated/sovereign cloud, with everything in the UK - including all the resources and necessary staff to fully support the operations - that would be more disturbing.
You're right though, I'd guess there's a lot of looking at the small print on contracts taking place right now!
To offer a slightly different perspective, if we look at Monday's incident with AWS, the fact is that the vast majority of people and businesses weren't affected, and probably wouldn't have known there was a problem if it wasn't for the excessive media coverage.
The Guardian, which seems desperate to make something out of nothing in its battle against the "Tech Bros", reports that "More than 2,000 companies worldwide were affected by this week’s outage, according to Downdetector, a site that monitors internet outages, with 8.1m reports of problems from users including 1.9m reports in the US, 1m in the UK and 418,000 in Australia."
Hang on......2,000 companies globally, and 8.1m problem reports? By any measure, that's pretty much peanuts when looked at in the grand scheme of things.
As The Guardian went on to report, "By Monday evening, Amazon said all of its cloud services had returned to normal operations".
By contrast, JLR was shut down for 5 weeks, and M&S a little longer.
Barely 48 hours after the AWS outage was first reported, for most people things are entirely back to normal.
That actually sounds pretty resilient to me.
TringSaint
Well-Known Member
The bank I know were deployed within a vpc in 2 AZ’s in UK with geo rep in Dublin using ec2 data centres, so public cloud but not public cloud if you see what I mean.
I’m not overly familiar with AWS terminology (I am apps, not tech), but as I understand it their sovereign cloud offering is independent from the public cloud data centres in the AZ’s, and has to be ordered specifically. EC2 then provides scalability on top of the AWS instances. I would expect both are likely to be administered using global resources, so would have connectivity to the US, which may be why they were affected.
All speculation on my part though.
TringSaint
Well-Known Member
Yeah, likely to be the case.
Unfortunately that may not have been clearly understood by the companies using non US data centres!
Certainly will raise some pointy questions!!
At the risk of sounding like the four Yorkshiremen sketch…..luxury!
My first foray into computing involved filling out punched cards and sending them off to Imperial College, getting them back the following week with (hopefully) some output. Progressing to teletype and magnetic tape felt like being on the command deck of the Starship Enterprise! After that it was the Research Machines 380Z, before moving on to IBM 8100 series running DPPX and their mainframe systems running CICS, both being programmed in COBOL. Then using Informix 4GL on HP, IBM and Tandem hardware, before the world of Client Server and then Cloud finally took over.
I still have a couple of the 8” floppy disks we used to cart around to install software in the mainframe days, together with the first Motorola mobile phones we used for remote support. I take these along to presentations I deliver at universities, to show students just how rapidly technology has evolved.
What is more frightening still is that the pace of technological evolution hasn’t stopped accelerating,
There is a great techie YT chap who reveals some of the truly scary advances in this evolution. The channel is Low Level Learning.
He presents complex hacks and code in relatively lay terms.
The following recent presentation is an AI discovered zero day...
That's a pretty interesting video, thank you.
I know of several companies that now rely on AI to generate source code, thereby eliminating the need for programmers. Using AI to detect flaws and anomalies will improve the overall quality of generated code, as well as helping in terms of validating legacy code.
It would have been interesting in the video to see them testing different LLM's against O3, to see if that improved the detection rate. Maybe that's something they will go on to look at.
Similar threads
