cloud data and machine learning

What exactly is Cloud Computing?

Cloud computing is the delivery of computing services like servers, storage, databases, networking, software, analytics and more over the Internet. In basic words, cloud computing is computing that is based totally on the web.

You need an internet connection to access the files. Facebook, Twitter, and Instagram are all examples of cloud-based applications.

Think of cloud as computers with unlimited storage with high processing speed located in specific areas around the globe. These are what we refer to as data centers.

Cloud service providers offer two types of cloud. Private and public. Private is only accessible within an organisation while public is accessible to the general public. Security is more enhanced in private than in public.

Three services model are based on cloud computing. These are:-

  1. Infrastructure as a service (IAAS) – This is access to resources offered by the provider . It can be virtual devices or physical devices e.g. virtual storage.
  2. Software as a service (SAAS) – Using software applications offered by a provider for an end user benefit, for example email services.
  3. Platform as a service (PAAS) – Using different environments offered by a provider to run your application, it can be in PHP, JS, Ruby or Python.

The cloud is analagous to the “fog of war” We could just as easily have called it “the hairball”.

Cloud computing is the use of services in a remote, untouchable location on demand.

On demand is important. You are renting a service. It is the data processing equivalent of visiting a brothel.

If it floats, flies, f@#ks, or floating point calculates, rent it by the hour.

Advantages.

  1. You can switch it off like a tap, and it is common for services to be charged by the second. This can be very useful when insolvency looms. In practical terms you can have thousands of cores for 10 minutes, rather than 10 cores for thusands of minutes. This is often cheaper than the hit to your power bill alone would have been, and this scaling can be called up on demand (in theory). You do not want to be paying for anything as a service unless you can shut off the cashflow on demand.
  2. Similarly your personal devices can be cheaper and “thinner”. I can have Windows10 on a 640MHz Pentium-M EeePC with 2GB RAM thorough the magic of “Remote Desktop”. If I break or lose it, my data is safe. CBP can’t get anything either, and because my disk is virtualised, I have at lease some protection from Government and LEO backdoors.
  3. Like most forms of outsourcing, there are few assets on the balance sheets. (ICANN matters may lead to de facto assets). No depreciation, and everything is tax deductible.
  4. This also means there are no property requirements for those assets. A migratory office community can jump between cheap broken lease properties at the drop of a hat. Ditto for fleeing the country.
  5. You can often bid for spot pricing, although in the extreme people that do not reserve a booking may end up paying more.
  6. It is often beyond the reach of law. Certainly it can be beyond the reach of consumption taxes such as VAT/GST if the service is provided overseas.
  7. Many cloud providers have seriously tuned infrastructure. The performance of the storage and processing is very hard to match on your own dollar.

Drawbacks

  1. You are putting your data and business at risk. You may not have equal standing in the crowd when things go wrong.
  2. If you lose data on the cloud, it is really lost. You can’t hold it in your hand. This is not like your parents passing and not recording their frequent flyer miles password, this can be way worse.
  3. The cloud is very expensive if you are not good with a calculator. A PC with 20TB of storage can be put together for $20,000. On the cloud this is a million dollars a year. Leave a few servers in the oven, and the bill shock may end it all for you.
  4. Exit strategies can be tough. You need to have a plan to end your relationship. You can walk 20TB across a room, or take it on a road trip across the country. Getting it back from the cloud can take weeks, months and involve serious labour costs. Kudos to AWS for the Snowball, it goes some of the way. Stealing your data and transfering it (at some cost to you) elsewhere on the same cloud provider is way faster.
  5. You cannot make mistakes with access and identity management (AIM). You need to have multiple administrators, and a good password policy, as in do not lose them, and have the ability to change them. You won’t have an axe hanging on the wall either.
  6. It can take a while for instances to launch, and resource starvation in some geographic areas can be a seasonal problem. In times of demand there are no guarantees.
  7. Employees with actual skills, rather than people who have just fallen into it because it is a trend, can command serious salaries. More so as things become more “automated”. Documentation can be critical in these cases. Secure this as if were cash.

In conclussion I would say, Cloud computing is changing the face of technology and I would recommend that everyone acquire skills in this area.

Leave a Reply

Your email address will not be published. Required fields are marked *

Bitnami