1.1. Introduction
When discussing the development of IT in recent years, one cannot overlook the share of Cloud solutions among others. Let's explore what cloud solutions, technologies, etc. really are.
Cloud computing (or cloud services) is a specific set of tools and methods for logistics, storage, and processing of data on remote computing resources, which include servers, storage systems, and data transmission systems.
When producing an IT product, be it a landing page, an online store, a high-load portal, or a database management system, there are at least two options for hosting your product.
On the client's premises (English — on-premise) or in the cloud (English — cloud). However, it is not possible to definitively say which option is more cost-effective in general.
If you are using a server hosting a small database that does not require fault tolerance and a simple website without significant load — then yes, on-premise hosting is your option. But as soon as your load and requirements increase, it is worth considering moving to the cloud.
1.2. Clouds Among Us
Before discussing how clouds are provided, it is important to understand that the story of clouds is not just about large giants in the IT sphere and their internal services; we use cloud computing every day.
Today, in 2019, it is hard to find someone who hasn’t used Instagram, email, maps, and traffic on their phone. And where is all this stored and processed? That's right!
Even if you, as an IT specialist in a company with at least a small branch network (to illustrate), set up a storage system in the infrastructure, no matter how you grant access to the resource, whether through a web interface, FTP, or Samba — for your users, this storage will be a cloud that is located… somewhere out there. Not to mention the familiar things we use at our fingertips dozens of times each day.
2.1. Types of Cloud Deployment
Okay, cloud. But it's not that simple. We all come to work — salespeople, IT specialists, managers. It's a broad concept; everyone has a role and a classification. The cloud service can also be divided into 4 types.
1.Public cloud is a platform that is publicly available to all users either for free or via a paid subscription. It is most often managed by a specific individual or organization. An example would be an aggregator portal for scientific articles.
2. Private cloud is the complete opposite of point 1. This is a platform that is closed to the public, often intended for a single company (or a company and partner organizations). Access is granted only by the system administrator to users. These may be internal services, such as an intranet, service desk (SD), CRM, etc. Typically, the owners of the cloud or segment take information security and business protection very seriously, as private clouds store information about sales, clients, strategic plans of companies, etc.
3. Community cloud can be described as a private cloud distributed among several firms that share similar tasks or interests. It is often used when there is a need to grant application resource usage rights to multiple people or divisions from different companies.
4. Hybrid cloud is a type of infrastructure that combines at least two types of deployment. The most common example is scaling a customer data center through the cloud. This is done to save costs when it is not possible to move to the cloud 100% or due to security and legal compliance issues.
2.2. Types of services
Great, the types of deployment are so different, but there must be something that unites them? Yes, these are the types of services, which are identical for all types of clouds. Let's take a look at the 3 most common.
IaaS (infrastructure as a service) Infrastructure as a Service (IaaS). In this option, you are provided with servers in the form of virtual machines (VMs), storage, and networking equipment, on which you can deploy your desired operating system, environment, services, etc. Although I am currently focusing heavily on the Yandex cloud, my introduction began with GCP (Google Cloud Platform), so I will provide examples based on it. An example of an IaaS solution in GCP would be the Compute Engine element. That is, it is a simple VM where you choose your operating system, handle software configuration, and deploy applications yourself. Let's consider an example. You are a Python developer and you want to create a website with a backend in the cloud, considering only the IaaS option. You need to take one VM on which the site will run. For this, you need to install the OS (which is chosen at the instance creation stage in GCP), update the package manager (why not), install the required version of Python, Nginx, etc. ... Create a fault-tolerant database cluster on three VMs (also manually). Consider logging, etc. This is inexpensive and time-consuming, but if you want maximum flexibility — this is your choice.
Next, closer to simplicity and cost is PaaS (Platform as a Service). Here, you also receive VMs, but without the ability to change the configuration so flexibly. You do not choose the OS, software set, etc.; you get a ready-made environment for your product. Returning to the same example, you buy two App Engine instances in GCP, one of which will serve as the database while the other acts as web servers. You do not need to configure serving programs; you can launch a production environment right out of the box. This costs more, as you must agree, labor has its price, and a whole Script has worked for you. However, you receive a ready-to-use platform.
The third main option, standing above the rest — SaaS (Software as a Service). You do not make fine-tuned configurations of VMs; you do not configure them at all. You do not need to be an IT specialist, you do not need to write code, and you do not need to create a backend. Everything is already prepared. These are ready-to-use, deployed solutions like GSuite (formerly Google Apps), DropBox, Office 365.
3.1. What is under the hood?
Got it in your head? Okay, let's move on. We purchased a VM, worked with it, destroyed it, and bought another 10. We don’t acquire hardware, but we know it must exist somewhere. When you integrated storage systems into your business infrastructure, you probably set them up in a rack in your server room. Cloud technology providers offer you a part of their server facilities, just on a much larger scale. These are known as data centers (DCs). They are massive complexes located almost all over the planet. They are typically built near places that can provide natural cooling for at least part of the year, although some might even be built in the deserts of Nevada. In addition to installing several hundred racks in their colossal hangars, providers also take care of heat exchange (everyone knows that computers shouldn’t be frozen or overheated, right?), and they prioritize the physical security of your data, making it really hard to access the DC illegally. However, the methods for data storage in DCs differ among various providers; some create distributed records across different data centers, while others store them securely in one location.
3.2. Clouds now and in retrospect. Providers
If you dig into history, the first prerequisites for creating today's cloud platforms date back to the mid-1970s during the development and implementation of the ARPANET internet prototype. At that time, it was said that someday people would be able to access all kinds of services over the network. Over time, connections became stable and moderately fast, and in 1999, the first commercial CRM system emerged, offered exclusively by subscription and representing the first SaaS, with instances stored in a single data center. Later, the company branched out into several divisions that offered PaaS by subscription, including a specific case of BDaaS (Database as a Service). In 2002, Amazon launched a service allowing information to be stored and processed, and in 2008, it unveiled a service where users could create their own virtual machines, marking the beginning of the era of large cloud technologies.
Nowadays, it's common to talk about the big three (although I see a big four emerging in six months): Amazon Web Services, Microsoft Azure, Google Cloud Platform… Yandex Cloud. Particularly nice for the latter, as it fills one with pride when compatriots swiftly break into the global arena.
There are also many companies, like Oracle and Alibaba, that have their own clouds but, due to certain circumstances, are not as popular among users. And of course, the guys—hosts, who are also providers offering PaaS or SaaS solutions.
3.3. Pricing and Grants
I won't dwell too much on the pricing policies of providers, as otherwise, it would turn into open advertising. I want to highlight that all major companies offer grants ranging from $200 to $700 for a year or shorter periods, so that you, as users, can test the power of their solutions and understand what you specifically need.
All companies from the big three… or soon to be four… also provide opportunities to join their partner ranks, conduct seminars and training, offer certification, and give discounts on their products.
Source: habr.com
