Let's Make Robots!

Cloud computing - the next world crisis?

I have just used my first internet based software tool and I found the number of times it could not save my changes because the server did not respond or gave a bad response to be very annoying!

When I first heard of cloud computing I thought it was a foolish idea. Now more than ever I think any company that becomes dependant on cloud computing is going to suffer badly.

Ok, this is just my personal opinion, maybe I'm wrong, here are the reasons I think it's a bad idea.

  • Your computers are always going to crash or have hardware failures. Moving to the cloud just means you will suffer when their equipment fails as well as when your equipment fails.
  • Security will always be a problem, even if a cloud server was 100% secure (never going to happen) then malware on your computer could still steal it when you access that information. Recently "Hackivist" have proven that if anything, your data will be less secure on a big companies server. 
  • You are now more dependant than ever on a good internet connection. In the short history of the internet the number of ways your service can be disrupted is steadily growing. Every thing from old ladies cutting the cable while scrounging for metal to natural disasters. Future sun activity is going to wreak havok on the satelite communications systems in the next ten years according to the latest reports.
  • Even if everything is working perfectly, when dealing with large files such as high resolution photos or maps then a local harddrive will always be faster than an internet connection. Especially as solid state harddrives come down in price.

At best, if security is not a big issue then I would only use cloud computing as a secondary backup method. I'm sure there are IT experts on LMR who know a lot more about this subject than I do. If I am wrong then please enlighten me because at this stage it seems as though businesses moving to cloud computing may be nothing more than a cost saving exercise.

 

Here is a link to a news story concerning future solar activity:

2013 - perfect solar storm

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

I agree OddBot. I, too, thought "cloud computing" was silly, --just another demonstration of corporate silliness.

[ --and the first word in demonstration is demon. The second word is demons, plural.]

However, I am fairly certain that it will spread as the internet did (mostly because of the recent growth of internet apps. and access by cell phones, etc.) and more and more functions will be available.

I thought the cloud computing would take the form of every home computer being part of the cloud (helping with calculations, etc.) but with the increased used of cell phones, which do not have that much memory nor processing power, I am not so sure about that part.

--------------------------------------------------------------------

While above the clouds, the sun is just a lazy quiet ball of gas...  ----NOT!

This is our Sun on a quiet day. recently it has been much worse, and GOODNESS....

              Seeing it in UltraViolet light, NOW we know what those odd sunspots really are !!!!

                        The ROBO-SUN.

     "OH NO !   LMR has Robotised the Sun."

                                             It's a "START (the day) HERE" robot.

 

 

For a big company there is always the possibility to build their on private cloud. As for public clouds, personally I think those are just the test beds for the private ones. The price of a private cloud may be on the same order with a "common" clustered virtual environment (I am talking about high availability clustering of virtual servers, redundant FC SAN's and networks) so for a big company having a private cloud gives advantages and disadvantages, but may be more a matter of public image - I use a cloud, I rock.

From an individual or small company point of view, moving to the cloud may have some advantages but I would move only what is absolutelly needed there. The cloud may offer you a higher availability at a lower price then running your somehow equivalent solution.

As for the Internet links, selecting two providers with different physical communication channels (at least one with optical fiber) gives a "safer" Internet link.

When it comes to digital signal processing (DSP) on a mobile device, processing a moderate amount of data in the cloud can make a stupid device smart- the cloud can be used as the mothership in a sense.  Also, access to the latest and greatest DSP algorithm running on a remote server by default is another advantage.  One example is google's voice processing for Android moblie operating system.

I also think cloud computing can make a low power device (something that lasts over one year on a coin cell battery) seemingly capable of complicated and otherwise processor-intensive tasks which typically consume a bit of power.

I think companies running on the cloud versus with a downloadable program gain some security from their software being reverse engineered.

That being said I have no personal interest using an online development environment to program for the reasons you posted. The web-based C++ development environment of the mbed platform just scares me. I have two smartphones and no data plan and I highly value local processing even with limited hardware.

Another thing I don't like about cloud services is they can change or be discontinued, so continued use is not gauranteed.  I like a program to run on my computer if I plan to use it for a long time.

I have used the mbed platform and while I have not had any issues with it, my fear is that one day the environment will cease to exist and then the chip is practically useless. Unlike arduino or mplab which I have on my hardrive.

The cloud is good as you pointed out for distributed intelligence but there's no way I would trust my data out there well certainly not without a local backup and that defeats the purpose? And I also feel as if the right to possession after paying is being eroded by this model.

Is the cloud encouraging poker machine or arcade game machine business models where you have your fun for a dollar and then you don't own anything after it? It seems to me that is where media companies would like to take the cloud concept for their businesses.

So to me the cloud offers only one positive but many negatives.

> may be nothing more than a cost saving exercise.

Yes. Yes it is. Not sure if you're frowning on that or not though.

One way to look at this: if I was starting a business right now there's no way I'd build infrastructure. I'd get as fast an internet connection as I could afford and put everything I needed in the cloud. Then let people have at it, with some kind of enlightened consumer device access policy. There's no point running mail or file servers these days unless you have a non-technical driver to do so (like, you're public sector, R&D, etc.) It's hard, very expensive, and so many other people do it better. Better to spend the money on growing your core business, not an IT shop.

For established businesses, moving to the cloud is trickier. Certainly, for "enterprise class" IT there are some big savings to be made in paring away IT until you basically end up with a specialist group managing vendor relationships, with the design and build bits of IT out in the business. It's going to be a very tough world for old school "IT department" people in the next 5-10 years.

Security's hard everywhere, and in a large IT shop it's a nightmare. An endless forced march of patching and checking. If it's in the cloud it tends to be somebody else's (expensive) problem. Connectivity, yes, fast is essential. That's just one "non func" of many though.

As to machines dying, yes, they do. What matters is your architecture - if your applications are written correctly, meaning resiliently, then they can survive that. If you're starting an application from scratch for the cloud then you shouldn't really be thinking in terms of machines; think services, reuse, scaling, etc.

Your last point's an interesting one, re large files. There are certainly cases where the cloud's not the place - image editing's probably the standard use case for that.

(20+ years in IT, just for the record)

Indeed cloud computing may be the future but it may cause more problems than is solves. Security is a huge issue, instead of millions of personal computers, information form millions of users will be in one spot, so cloud protection will always be an issue this is why I do not think it will catch on.

I only frown on the cost cutting aspect because it is usually shortsighted and could lead to much greater cost in the long run.

It's interesting that you skipped over the part about how you run your new business when an old lady has cut through the cable, an earthquake has damaged a cable or solar activity has interrupted the satelites.

I admit, last time I worked in IT we were upgrading over 300 computers from Windows 3.11 to Windows 95. Printer repair was my normal job at the time.

 

I wasn't trying to do a line-by-line, but to address your point I doubt it matters if your services are hosted off site or on if you lose your link. Companies who need modern IT services pretty much by definition can't live without the Internet, as you can't do business without it, and if you're thinking about the cloud then you're no mom-and-pop shop paying by cheque.

Indeed, cloud computing offer great business opportunities and more flexibility to the user. There is a question of privacy and security, the risks seem higher, I am currently trying some cloud services provided by our San Diego data centers and they work amazingly for now.

I agree, this is why the next world crisis will be internet related.