The End of the Cloud Is Coming

Megalith

Megalith

24-bit/48kHz
Staff member
Joined
Aug 20, 2006
Messages
13,000
It’s a bold statement, but one author argues that we’re facing the end of the cloud for four reasons: it can’t meet long-term scaling requirements, is centralized and vulnerable, demands trust but offers no guarantees, and makes us — and our data — sitting ducks. He suggests that the best alternative is peer-to-peer web technologies.

Building a scalable, reliable, highly available web application, even in the cloud, is pretty difficult. And if you do it right and make your app a huge success, the scale will cost you both money and effort. Even if your business is really successful, you eventually hit the limits of what the cloud, the web itself can do: The compute speed and storage capacity of computers are growing faster than the bandwidth of the networks.
 
No, not a chance.

Small business running one Windows SBS server for storage, AD/DC and Exchange - That's vulnerable, and laughable.
 
49nocloud.jpg
 
I read this: "It’s centralized and vulnerable. The other issue with centrally storing our data and programs is availability and permanence. What if Amazon’s data center gets flooded, hit by an asteroid, or destroyed by a tornado? Or, less drastically, what if it loses power for a while? The data stored on its machines now can’t be accessed temporarily or even gets lost permanently."

At this point I started scratching my head wondering if this guy really knows what the cloud really is.


See amazon's infrastructure of their AWS sites:
https://aws.amazon.com/about-aws/global-infrastructure/


Sounds like the guy is trying to come up with a concept for data to be encrypted, associated to name, and somehow be decentralized. I foresee that most devices become more like thin clients so nobody loses data. Being able to switch phones all day long and never worry about downloading or losing data because it is all on the cloud.


Also noticed at the bottom "Viktor Charypar is a Tech Lead at UK-based digital consultancy Red Badger."


again scratching my head.
 
Whew, that's some uninformed opinion right there. I do agree with the "demands trust but offers no guarantees" - you are at the mercy of your cloud service provider. Microsoft, however, has the Azure Stack now, which allows organizations to host azure assets locally.
 
At work we are migrating A LOT of shit to the cloud. Email, applications, personal drives etc.

The email side of things has been a real pain in the ass for a lot of folks from losing permissions to general slowness...

I hope things get better in the long term, short term its made things pretty fucking busy for the dept that I am in.
 
dR.Jester said:
At work we are migrating A LOT of shit to the cloud. Email, applications, personal drives etc.

The email side of things has been a real pain in the ass for a lot of folks from losing permissions to general slowness...

I hope things get better in the long term, short term its made things pretty fucking busy for the dept that I am in.
Click to expand...

Teething issues on such a massive switch are to be expected, eventually all problems will be sorted and you'll be fine.

I have to say, in terms of migration, I find G-Suite to be better than Office 365.
 
actually he may be right, companies do not want to invest into protecting their clients data. they do however want to keep getting paid more and more for their services.
unfortunately the companies do not want to pay for knowledgeable support staff nor do they want to pay for continuous R&D . so the clouds ends up being a terrible platform for business and their customers.
these companies do not want to pay the price to keep their product up to date, and resist making changes to their primary code base. Companies do not care about your data nor do they care about your accessibility to it.

while the cloud is a great idea and if companies had the balls and willingness to actually support the platform while letting the customer control their data, while ensuring it is protected , the cloud could be a successful platform .
However in reality all the cloud turned out to be, was a data gathering platform enabling providers an alternate revenue stream for selling their clientel's data to third party sources.

i will not be surprised to see the return to local iron as the cloud has failed to pretty much deliver on its hype.
 
A lousy IT staff will result is an unsecure Cloud, just like they will result in an unsecured local network.

Other than some file sharing, there is no benefit on moving any of our office applications to the cloud. Our costs would be higher and I'd have to add more bandwidth to our internet connections resulting in even more costs.
 
nutzo said:
A lousy IT staff will result is an unsecure Cloud, just like they will result in an unsecured local network.

Other than some file sharing, there is no benefit on moving any of our office applications to the cloud. Our costs would be higher and I'd have to add more bandwidth to our internet connections resulting in even more costs.
Click to expand...

You're costs would be higher than Windows licensing/IT staff wages?
 
people going cloud are idiots. When the US infrastructure matches that of say, South Korea; then talk.
 
beemanit said:
I read this: "It’s centralized and vulnerable. The other issue with centrally storing our data and programs is availability and permanence. What if Amazon’s data center gets flooded, hit by an asteroid, or destroyed by a tornado? Or, less drastically, what if it loses power for a while? The data stored on its machines now can’t be accessed temporarily or even gets lost permanently."

At this point I started scratching my head wondering if this guy really knows what the cloud really is.
Click to expand...

Probably read some article about someone losing everything because a hard drive faulted at CrashPlan.com and just assumed that's how it all works.
 
lilbabycat said:
people going cloud are idiots. When the US infrastructure matches that of say, South Korea; then talk.
Click to expand...

If we ever reach that level of population density, the planet will be done for. There better be cheap rates to Mars by then.
 
BulletDust said:
You're costs would be higher than Windows licensing/IT staff wages?
Click to expand...

We might be an exception since we are a Microsoft developer.
Microsoft gives us a boat load of free enterprise licenses for production system so our main costs are for the hardware. :D
Also I am the IT department, and they would still need me to support stuff even if it was in the Cloud.

Biggest challenge is the massive amount of VM's used for development, testing and demos.
50+ TB of VM's would be very expensive in the Cloud. The VM's are also why even a Cloud based backup doesn't make sense.
I'd also need to upgrade the office internet connections to gigabit to support all the traffic.

So yes, the cloud WOULD be more expensive.
 
It's also expensive. At least 2x for us over hosting ourselves, plus the time spent completely changing the way we do IT.
 
bigdogchris said:
It's also expensive. At least 2x for us over hosting ourselves, plus the time spent completely changing the way we do IT.
Click to expand...

The ones criticizing local hosting are probably the ones who have nice huge IT budgets and great bandwidth. When you work for a company that can afford only a single IT person and maybe 10Mb/1Mb internet( not sure what we have now, just upgraded to DSL from a 1/3 share of a T1), you may not have good results from cloud technology. Plus when you still have equipment worth $300k that is stuck on WinXP but still needs access to the Laboratory Information System, you probably don't want to host that out on the cloud.
 
jedijeb13 said:
The ones criticizing local hosting are probably the ones who have nice huge IT budgets and great bandwidth. When you work for a company that can afford only a single IT person and maybe 10Mb/1Mb internet( not sure what we have now, just upgraded to DSL from a 1/3 share of a T1), you may not have good results from cloud technology. Plus when you still have equipment worth $300k that is stuck on WinXP but still needs access to the Laboratory Information System, you probably don't want to host that out on the cloud.
Click to expand...

This is the thing. Cloud based services are perfect for small business of around 5 - 20 employees/terminals with no real budget set aside for IT support staff.

In such situations a Microsoft based SBS is a complete waste of money.
 
nutzo said:
Biggest challenge is the massive amount of VM's used for development, testing and demos.
50+ TB of VM's would be very expensive in the Cloud. The VM's are also why even a Cloud based backup doesn't make sense.
I'd also need to upgrade the office internet connections to gigabit to support all the traffic.

So yes, the cloud WOULD be more expensive.
Click to expand...

Actually - you should take a look at how VMs are managed in the various clouds because it is rare that you would "lift and shift" your entire VM collection and data and persist it in the same way as you do in local hosting. What I mean specifically is that your VMs are not running 24/7 for the cases you outlined. For development and demo scenarios, spinning up VMs when you need them is a flexibility offered by cloud at a speed which it is rare that a local hosting could match (minutes/hours versus days or weeks). Additionally, think about the benefit of templates and available software/images (from the services themselves to ones you make and store, etc.) in a cloud environment - your clients aren't doing the downloading and uploading from their workstations anymore. Your regular traffic beyond the initial upload/migration would not necessarily need gigabit. All of this generally adds up to be a lower operational cost than local hosting or private hosting.

I've found that cloud entails looking at how datacenters and servers are managed a bit differently from the traditional sense.
 
g-money said:
Actually - you should take a look at how VMs are managed in the various clouds because it is rare that you would "lift and shift" your entire VM collection and data and persist it in the same way as you do in local hosting. What I mean specifically is that your VMs are not running 24/7 for the cases you outlined. For development and demo scenarios, spinning up VMs when you need them is a flexibility offered by cloud at a speed which it is rare that a local hosting could match (minutes/hours versus days or weeks). Additionally, think about the benefit of templates and available software/images (from the services themselves to ones you make and store, etc.) in a cloud environment - your clients aren't doing the downloading and uploading from their workstations anymore. Your regular traffic beyond the initial upload/migration would not necessarily need gigabit. All of this generally adds up to be a lower operational cost than local hosting or private hosting.

I've found that cloud entails looking at how datacenters and servers are managed a bit differently from the traditional sense.
Click to expand...

Would never work in our environment. Most VM's are not left running, but are only spun up as needed.
Demos are usually run locally on laptops. Latest VM is copied down (on the gb network) in the office as needed.
We try running demos over the internet, at least half the time they end up running it locally due to internet speed/access issued at the customer's site.
Rule is never try to run a demo at a customer location unless you also have the latest image locally on your laptop.

We are constantly spinning up new VM's (from a base standard image) but then install the latest version of our software and other numerous vendors software we integrate to.
Having to constantly upload all this software to the cloud would take way more time.
Uploading/downloading VM's would eat up more internet bandwidth than I currently have available.

The more data you are supporting, the less useful the cloud is.
 
The cloud is just like anything else in IT, not a one size fits all solution. Different companies, different goals, different resources, all that has to come in to play when making a decision for the company's best interest.

Sometimes some services make sense to move to the cloud, sometimes they don't.
 
For small and medium businesses, you are fine.. for large businesses the cloud isn't all that great. They are better served using the cloud for SOME things but private cloud and onsite for the rest.
 
You're always going to run into the issue of the people paying for IT not knowing anything about IT.

That's how "the cloud" started. That's how it currently has issues.

The problem is not any technology though, just dumb or cheap company executives.
 
BulletDust said:
No, not a chance.

Small business running one Windows SBS server for storage, AD/DC and Exchange - That's vulnerable, and laughable.
Click to expand...

To me, cloud computing has always been a double edged sword. It's got it's place and makes sense for certain situations and not others.
 
  • Like
Reactions: raz-0
like this
How is distributed or peer to peer different from “cloud”?

My caveman understanding was cloud was anything out in the internet (not local machine or local network server)— if it’s out there somewhere in the pipes of the internet, it’s “in the cloud”

Doesn’t matter if that’s a single source server, a globally distributed repository system, or chunked up into little tiny bits and shared across thousands of computers — that part is all transparent. The only distinction in my mind was local, or internet.

Maybe I was wrong about clouds this whole time.
 
Bullshit. You cannot get close to the R&D and Speed for cloud, and it’s speeding up. Lose economies of scale and that goes out the window, you better believe that is what the ceo wants, he doesn’t give a shit about the technology. Just what it can do for him.

It is also generally more secure and more resilient than all but the largest enterprises can create. When you talk aws or azure misconfiguration or poor architecture causes the vast majority of problems. Think about it, can you imagine just how much amazon spend on security for aws. It would bend your mind and you can’t match it. Idiots can break it, as they always could though.
 
Define cloud.

I thought he was going to say it was replaced with mainframes...
 
AWS is so cheap, it's not going away. I don't like the 'cloud' concept. It reaks of lack of accountability for my personal use. But I can see where if you needed a bunch of servers for a year only like some event or something. The VM thing running on AWS or similar is the way to go. To buy or rent hardware and staff for just a year sounds ludicrous in comparison.
 
jpm100 said:
AWS is so cheap, it's not going away. I don't like the 'cloud' concept. It reaks of lack of accountability for my personal use. But I can see where if you needed a bunch of servers for a year only like some event or something. The VM thing running on AWS or similar is the way to go. To buy or rent hardware and staff for just a year sounds ludicrous in comparison.
Click to expand...

But AWS is only one type of cloud. There are many different types including "private" clouds that may be hosted on premise. AWS is not the end all solution in cloud and neither is google or microsofts offering. Unfortunately TFA completely misses the point of cloud computing.
 
Brian_B said:
How is distributed or peer to peer different from “cloud”?
Click to expand...

What he is talking about is known informally as "grid" computing and it's been studied at Palo Alto and IBM for quite a long time now (since early 2000s). I know because I was involved at one point and it was very promising. The problem or rather the limiting factor was IPv6 (with security/privacy as secondary concerns as always) implementation which remains, let's say, not quite there. Among the advantages of grid would be geobalancing which effectively is tying performance (caching) and geographical location together for something like, say, YouTube videos, which would reduce the load on central servers and hubs a ridiculous amount (which is a massive problem even with cloud dispersion), and just be far more efficient on the whole. (as an example, imagine your next-door neighbor had just watched a certain YT video that you are now requesting; it would fill it right from his cache to your device. now imagine that same concept done with mobile devices and in a many-to-one sharing structure) Unfortunately when I first got into the field during my university years (again early 2000s, now I feel old), it was still "twenty years away" and that's now looking optimistic. I got into cloud computing around the same time and I have to agree with him that it's an incomplete solution.
 
Last edited:
Sounds more like arguments against the "cloud" than real reasons "cloud" will go away. I really just hope the term "cloud" goes away. It's chintzy, cheap, and means nothing.

That being said, I have been skeptical of hosted services for some time. I have seen one downside recently - our company merged with another, and now have offices nationwide, and everything is now "cloud." And slow. Slow. Slow.
 
jardows said:
Sounds more like arguments against the "cloud" than real reasons "cloud" will go away. I really just hope the term "cloud" goes away. It's chintzy, cheap, and means nothing.

That being said, I have been skeptical of hosted services for some time. I have seen one downside recently - our company merged with another, and now have offices nationwide, and everything is now "cloud." And slow. Slow. Slow.
Click to expand...


Then they did it wrong.
 
So peer to peer we shouldn't trust cloud providers but having pieces of my data stored across a series of devices is OK?

To the cloud is slow. I would check your internet connection and network infrastructure before I blamed the cloud.
 
DeChache said:
So peer to peer we shouldn't trust cloud providers but having pieces of my data stored across a series of devices is OK?

To the cloud is slow. I would check your internet connection and network infrastructure before I blamed the cloud.
Click to expand...
The whole system was poorly planned. Cloud = saving money so instant migration to cloud resources. It's been 3 years since the merger and migrations, and we still maintain different accounts on each system. I know it is not the underlying "cloud" technology, but the full implementation. But we would have done much better to stay out of the "cloud."

"The Cloud" offers a technology Utopia if you just switch all your systems to "The Cloud," but without proper implementation (which the company I work for will never spend the money on) it just creates more headaches.
 
jardows said:
"The Cloud" offers a technology Utopia if you just switch all your systems to "The Cloud," but without proper implementation (which the company I work for will never spend the money on) it just creates more headaches.
Click to expand...

This is true of any technology.
 
You must log in or register to reply here.
Back
Top