IaaS (infrastructure as a service) for small and medium businesses

Grandpa IaaS was the Amazon EC2 (Elastic Compute Cloud) product, which established de facto many standards for the use of already ready-made computer resources on demand. In particular, APIs are growing in popularity when deploying and managing public cloud resources using Amazon EC2.

At the same time, many IaaS providers come to the market, differing in service level, hypervisor platform, management capabilities and other additional services. The growing number of options means that IT managers in midsize and large enterprises should consider not only the cost of computing, but also the service level agreement, service levels, and technical compatibility with the virtualization tools they use on their own site^[1]

As a result of conversations with manufacturers and their customers, it becomes clear that the IaaS market is entering a phase in which buyers are mainly pioneers. So if your company studied IaaS, but did not take a decisive step, then you are far from alone.

When IT managers consider using IaaS as a strategic platform on which applications are developed and deployed, a number of technical factors need to be addressed. For example, what hypervisor platform does the provider offer? VMware vSphere virtualization platform is used to provide many additional services. Sometimes they offer open source Xen, kernel virtualization mode (KVM) of Red Hat Corporation and other options.

In any case, the question arises how the service provider manages virtual machines and tasks solved using them. What storage options are available? How do IT professionals access failed systems?

There are also related issues, some of which are technical in nature, others not. They also need to be understood. How is the service level measured and what steps are taken to ensure rapid recovery in the event of an unexpected failure? What management tools will work in the IaaS environment? Can traditional host services be used in conjunction with IaaS?

What are the ways to manage costs? What are the provider's IaaS plans for disaster recovery? How can I access applications if I stop servicing? What will be the consequences for sensitive and regulated data when it is encrypted, accidentally lost, and litigated in the case of searching for and learning information that is used or stored outside the direct control of your business?

If part of the functions of the data center of the company is transferred to the IaaS provider, then the IT department will become something like an elephant in the office. It is understood that the widespread use of outsourcing will reduce the need for employees involved in cable laying and server rack maintenance. However, due to the migration of some tasks to the cloud, the number of experts on various issues - from storage and networks to system administration and security - should also decrease.

Problem Number 1

Problem number 1 - hooligan users. IaaS focuses on the management of virtual machines, so the risks here are slightly different from those inherent in other types of cloud services. The main risk is hooligan or illegal management of services. Operation of the IaaS requires monitoring administration and use. It is recommended that you use enterprise frameworks to manage services that help prevent employees from improperly accessing information or misusing services. It also helps prevent increased costs for virtual machines or personal use of resources paid by the company.

Cost Reduction and Competitive Advantage

According to the IT services provider Terremark Corporation, enterprises that have already outsourced IT, and especially those companies that widely use virtualization on their own site, are the most prepared for the transition to IaaS.

John Ingates, chief technology officer at Rackspace, said: 'IaaS, being a product provided on demand and paid in proportion to the scale of its use, moves the company to the cloud with the applications that it has today. This forces enterprises to move towards cloud architecture in order to move to the 'platform as service' service [platform as a service, PaaS] '.

Going back a step, we see that server virtualization in the company's own data centers has led to a tremendous increase in their performance when working with applications for x86 architecture. IaaS is the next step in the same direction. The equipment needed to solve new mobile problems using virtualization can be concentrated at the service provider, saving the company from the current costs of maintaining a private data center.

In addition to eliminating hardware, network configuration and maintenance, storage, security, and hypervisor platform costs, the IaaS sets the standard for delivering development and test environments quickly.

API and Customer Requirements Service providers most often use APIs to enable customers to configure and manage tasks in the public cloud. The Amazon EC2 interface has become a de facto standard. Despite this, large and small cloud providers provide access to managed systems through their own Application Programming Interface (API).

According to Ingates of Rackspace, the open cloud computing platform OpenStack enlisted the support of George Williams, who now holds the position of chief architect in their company. The resulting Rackspace Cloud Files API and others become standalone standards.

IT managers should not only know how APIs are used in public clouds, but also be familiar with the work done by the Open Data Center Alliance (ODCA) in June 2011. This group of users released eight sets of customer requirements that make up the open data center usage model.

In early November, ODCA and the Distributed Management Task Force (DMTF) group created by manufacturers announced a cooperation agreement to accelerate the development of an industry standard for cloud management. DMTF is credited with simplifying the routine operations of IT managers. Together, the two organizations hope to make life easier for IT managers moving to IaaS.

IaaS (infrastructure as a service) for small and medium businesses

Pros IaaS:

• Cost. Cloud applications are generally cheaper because they can be reused and share resources. No pre-investment is required, and there is no need for highly qualified staff.
• Scalability. The resources of a single server may not be enough, but cloud systems scale very easily, and easily cope with sharp changes in load.
• Reliability. Some redundancy of cloud systems means that they are significantly more reliable than autonomous systems.
• Performance. Cloud services are constantly monitored and improve performance.
• Safety. Cloud systems have a higher level of security than their autonomous counterparts - and above all, this is due to the existing customer concern about this issue.
• Access to higher-ranking infrastructure capabilities. Small and medium-sized businesses can use applications and services originally created for large enterprises (Salesforce, SAP, Siebel, Oracle).

Cons of IaaS:

• Binding to the Internet. Users must have a permanent connection - wired and wireless - to the Internet to access data and applications.
• Data control. Small businesses depend on their cloud service provider to protect their sensitive data. How do I make sure your data is isolated from competitors if you don't have your own server in the cloud? How do I control users' access to sensitive information and ensure data integrity?
• Shutdowns. Often, SLAs are unable to stem the devastating impact of outages on small businesses. Even a small simple can do enormous losses on the scale of small businesses.
• Mobile Device Management: You still face security challenges on your part because users continue to use personal devices (laptops, tablets, smartphones) to work online. How do I protect their mobile devices?

These are the main pros and cons. Depending on the specific offer and over time, this list may change slightly, but in the foreseeable future, the main points will remain exactly in this form^[2]

So, looking at the advantages and disadvantages of cloud structures, let's take a look at some of the things you need to consider before you transfer your IT infrastructure to a cloud service provider.

Own data centers

Does your provider use the services of other providers of their own services or use exclusively their own resources? This question is especially relevant if you are going to contact a regional provider. Does your provider rely solely on its own data centers? If not, you will inevitably face a lot of unsolvable problems.

Control

You put direct control of your servers and applications in the wrong hands. You're gonna have to trust your provider. By passing the management of your IT infrastructure to a cloud service provider, you give it the keys to the survival of your business. Choose a provider wisely.

Motivation

Determine for yourself why you are moving your infrastructure to the cloud, what you expect to get from it, and what you need to do.

Licensing

Do licensing terms allow your software to be migrated to your provider's cloud servers? Find out before you start the transfer process.

Geographical location

Provider IaaS - How to choose the right? Where is the cloud server you rent physically located? If your data is stored at the other end of the planet, and you may need physical access to them, it makes sense to look for a provider closer to home. Let me remind you that the law obliging to store personal data of citizens of the Russian Federation in Russia has already been signed - and will enter into force on September 1, 2016. How this will affect the cloud services market is difficult to say, but for medium and small Russian businesses, it clearly makes sense to think about choosing a cloud provider with datacenters in Russia. Just so that you don't have to change the provider in a hurry.

Total Cost of Ownership and Cloud Provider Prices

This is probably one of the most important items to consider. Do you benefit financially if you abandon your own hardware and switch to cloud infrastructure? In this regard, everything is determined by your needs and the amounts that you are ready to spend.

Service Level Agreement () SLA

Carefully review the guarantees your future provider provides regarding the availability and security of your data. There are cases when customers of the same Amazon lost their data as a result of a system failure - and could not do anything about it, carefully reading into their contracts. Unfortunately - after the fact.

Change of provider

What happens to your data if you decide to leave the provider? What will happen to the posted information?

Billing

How often will invoices be submitted, can you count on discounts and how will additional costs be paid? How exactly will additional bills be charged for peak load periods?

Technical support

How is the technical support procedure organized? Special attention - emergency calls in emergency situations, communication methods and guaranteed response time. It is unlikely that you will like if the server falls to write a message through the address form on the site and send it to the void, receiving an automatic response "Thank you, we will contact you within a week."

Proprietary Platforms vs Open Standards

By choosing a provider whose platform is based on open standards (OpenStack or Linux KVM), you get significantly more flexibility for the future. As the company grows, you will inevitably be faced with the need for further growth and development of your infrastructure. If you are initially tied to a single vendor using a proprietary platform (for example, VMware vSphere), you will not be able to migrate applications or data to a more appropriate infrastructure - or migration will be so expensive that it will simply bury your budget.

Cloud safety

Cloud-Based Another very important question. On the one hand, you must unconditionally trust your provider, which receives full control of all your business-critical data. On the other hand - as a rule, cloud providers offer a lot higher level of security than you could secure yourself. They are simply forced to track the most current trends in this area and use the most effective practices - without this, the industry would have been simply buried very quickly.

Data Retention and Recovery

What happens if your data is corrupted? Does your provider offer a reliable backup and recovery system? Will you be able to roll your data back to the known operational state of the previous day? Today, memory is cheap enough for files to have at least three backpacks in geographically distributed datacenters.

Evolution of IaaS

Infrastructure as a service (Infrastructure-as-a-Service, IaaS) has evolved from an exotic service to a mass service. Read in the joint project TAdviser and Selectel ' Evolution IaaS':

• What information technologies have made IaaS possible - as it exists today?
• What event can be considered a reference point in the appearance of virtual machines?
• What includes IaaS today and what will happen tomorrow?

Notes