The Business Professionals Guide to Understanding the Cloud

Author: Mark Bogan

This article is designed to be a to-the-point, easy-to-understand, and generally painless guide to the Cloud written for the business professionals. It describes the “What” not the “How” of Cloud computing. It will cover all the key features that make up the Cloud in 2018 and a good deal about how it’s being used. The only time you will see an acronym is when it is next to the actual spelled out words. It is my intent that you come away with a solid understanding of today’s Cloud and all that it has to offer.

It’s been almost twenty years since “The Cloud” came into being, yet to many business decision makers this elusive technology “still” remains firmly obscured in the fog. Ask any executive wandering the halls to explain the Cloud and you will surely receive a high energy round of sales and marketing slogans followed by the how amazing and ground breaking the technology has been for business. However, the chances are that you won’t hear much reflection on what the Cloud actually is or does. This phenomenon seems to trace its roots back to when the first executive innocently asked their dependable CTO to explain the Cloud. Of course, the rest is well defined history as the poor unsuspecting CEO was subsequently bombarded with a litany of acronyms, unintelligible English-like language and shocking animated gestures. Our CEO subsequently passed the word around “better that you don’t ask.” Fast forward nearly twenty years and we have a feature rich next-generation Cloud and, a vast community of decision makers often signing very large checks with very limited knowledge of exactly what it is they are buying.

As with studying most large topics it is best to start at the beginning, so in this case we shall start with what most are likely familiar with, the Clients of the Cloud, after which we will work are way into and through the guts of the beast. The Cloud, is essentially a virtual data center capable of being in many places throughout the world at once. As such, it is accessible by computing devices effectively everywhere all the time. This total availability capability of the Cloud has enabled existing client computing devices to improve in what they can offer and has given rise to a legion of new devices and potential new device uses. Today’s client devices also referred to as end-point-devices group into several general categories, End User, Smart and Thing.

The “End User” category include desktop computers, laptops and mobile devices. In addition to being able to access the Cloud directly for resources such as mail, files and applications, these systems also provide access to Virtual Desktop Infrastructure (VDI). VDI streams a configured virtual desktop over a network to a customer's endpoint device and controls access to targeted centralized information resources (Your company’s data) which of course these days is often also stored in the Cloud. Additionally, the world of client based applications that are now being made available as Cloud Hosted Services (CHS) are making working out in the field in nearly any profession a truly viable full-time proposition.

Some of the biggest ticket items being pushed and developed today fall into a different category called “Smart.” They include Smart Cities which are more and more able to bind their civil services such as traffic management, security, life-critical, first responder and video awareness to backend systems available everywhere via Cloud service technology. Take the AWS Smart Signals for example, Miovision TrafficLink monitors and manages traffic signals remotely to prioritize resources and solve issues before they escalate. An additional Smart example is Google’s WAYMO self-driving car technology. This project takes full advantage of the concept of the virtual data center everywhere all the time.

The last category is “Thing”, now being commonly referred to as the Internet of Things (IoT). This is a rather large group of devices such as thermostats, cars, lights, refrigerators, wearables, security etc. The IoT is effectively billions of devices going about their day to day functions communicating with each other potentially without any human interaction whatsoever.

Now that we have a general handle on a billion plus items able and willing to communicate with the Cloud, it is imperative to understand the key ways that all those devices are able to talk to the Cloud. On the high level of excitement lately is the new IEEE Standards Association (WLANs) on the 5 GHz band 802.11ac standard, Wait! What is that? Let me try again, in more simpler terms. it is better known as “5G” or “fifth generation”. Essentially this is the new standard of WiFi that we will all be using everywhere from Starbucks to Verizon. It is defined as a high-throughput Wireless Local Area Network (WLAN).

Once we have a wireless connection to one of our devices, there are several paths that we might travel to get connected to our Cloud. The first, and easiest is sending our connection through our 5G Wireless Access Point (WAP) and then jumping on the Public Internet directly and going straight to the Cloud; simple! And this is how the general public interfaces with the Cloud. However, in today’s every expanding world of devices (Think manufacturing for example), in some cases we will need more control over how and what data and what level of data is sent to the Cloud. Here, two key technologies come into play, the first is referred to as “Edge Computing”. Edge Computing is used to help in running the Internet of Things (IoT).

Edge computer/devices serve as the local communication points for that billion plus Internet of Things running around out there in the world. Effectively Edge computing places additional devices at a network’s outer barriers. An edge computer often serves as a data collection site, rolls up whatever given data is submitted for analytics and then submits an at least partially transformed version of that data to the Cloud for storage and processing. It is said that an Edge server reduces the “Backhaul” of data, that is, it reduces the amount of data that needs to be sent across the wire and stored at the central repository in the Cloud. A simple conceptual example would be the Cable companies that reduce their Backhaul by forwarding non-live audio and video material to edge distribution points at the major broadcast news organizations for broadcast in the evening or ongoing news.

Edge computing is very powerful but requires special edge devices to be located in the remote locations. FOG (our second IoT supporting technology) computing however, leverages the processing power on existing computers, bringing into the fold a continuum of processing power that can process IoT data closer to where it is produced. An example of a FOG solution would be a set of water treatment facilities that uses the Cloud to control all operations. However, traditional Cloud computing causes large round-trip delays because of the enormous amounts of data generated from the distributed data sources that have to be transferred back to the Cloud for processing (The backhaul). Cloud computing by itself cannot meet the real-time requirement of low latency in early warning applications such as water quality safety and human safety, therefore computing devices are needed closer to each treatment facilities operational functions. FOG allows for the necessary improved speed in calculation response time that enables those critical real-time functions to occur.

Lastly, there are times when we simply need a secure connection that is always available for one or more computers to have access to our Cloud services. In this case what is called for is a site to site Virtual Private Network (VPN). Often a VPN is needed to link our non-Cloud based private data centers and our offices of work to our Cloud facilities in a standardized reliable and secure manor. In this case, the VPN gives us just what is needed.

So, we have made it to the front door of the Cloud and now it’s time to go inside and see what it’s all about. However, in order to get access, we need to prove who we are by providing our assigned credentials to the Identity Access Management (IAM) sub system. IAM has been getting a lot of attention lately, it is a framework that facilitates the management of digital credentials throughout an enterprise. Put more simply it is a combination of enterprise application security and the associated analytics that enables the right individuals to access the right resources at the right times and for the right reasons. Additionally, it encompasses the necessary reporting to understand who has access to what, where and when.

Once authenticated by IAM, we will then get access to the Cloud’s Metered (aaS) as-a-Service modules. Metered services are effectively the way Cloud Service Providers (CSP) generate revenue. The concept is sometimes referred to as "on-demand” and effectively allows you to rent a given service based on a pay-as-you-go model. At the top of the list of Cloud features that we are interested in are the Software as a Service (SaaS) applications. SaaS breaks down into two general categories, Cloud Provided and 3rd Party. The common application that are Cloud provided tend to be applications that all businesses and or business applications make use of. A couple of examples should make this clear. The first is Database as a Service (DbaaS) which allows users to access a given database that is loaded into the Cloud eliminating the need for setting up physical hardware, installing software or configuring for performance. Another example would be Analytics as a Service (AaaS), think Google Analytics only targeted at the Cloud service itself. How about Data as a Service (DaaS)? Now you won’t even have to obtain transform, slice or dice a common dataset, you can just rent it pre-prepared and use it when you need it. And then there is Insights as a Service (IaaS), this allows you spot trends from pre-configured data analytics.

More and more 3rd Party Cloud based services are being designed to run in the Cloud under the metered model. Typical examples are Customer Relationship Management (CRM), Enterprise Resource Planning (ERP), Product Lifecycle Management (PLM), Supply Chain Management (SCM), Supplier Relationship Management (SRM), and Accounting. In short, just about every application vendor out there has released or is trying to release a Cloud-based version of their application to capitalize on the metered pay-as-you-go model.

The other types of metered Cloud services target operating system and infrastructure resources. First, Desktop as a Service (DaaS) streams virtual desktops (discussed previously) over a network to a customer's endpoint devices allowing access via desktop software or a web browser. All the necessary support infrastructure, including storage and network resources, lives in the Cloud. The second type is Platform as a Service (PaaS) which provides a platform allowing customers to “develop,” run, and manage applications without the complexity of building and maintaining the development environment infrastructure. Lastly, we have Infrastructure as a service (IaaS).

IaaS provides the virtual computer operating system instances that you can log into to run whatever it is that you want to run in the Cloud. It also provides a common set of application support services that include “Big Data” which is a storage infrastructure that is designed specifically to store, manage and retrieve massive amounts of data, it is often referred to as NoSQL and is based on “non-tabular” data structures to hold and access information. Also, Relational Database Services (RDS) the typical “tabular” Structured Query Language (SQL) based databases that run the more typical business world applications with which you are most likely familiar. Additionally, Message Queuing which is often referred to as a “message bus” and allows data to be sent through queues (FIFO) or topics (Broadcast) from one application service to application service(s).

Finally, the IaaS includes detailed run-time Cloud services such as, billing (what you owe for aaS services), Cloud monitoring (how your infrastructure resources are performing), Cloud logging (exceptions and other events that may occur) as well as other key infrastructure services such as load balancing (think web farm, where traffic can be serviced by multiple virtual server instances) and clustering (a collection of servers that communicate with each other and provide redundancy to each other to make a set of services highly available to clients), as well as storage resiliency, such as archiving and recovery.

Now it’s time to get down to where the real action happens in the Cloud. Resource abstraction is truly at the heart of what gives a Cloud its value proposition. The notion of a Cloud is based on the concept that the actual physical Cloud resources are shared by all the Cloud’s consuming customers. This offers serious advantages to all parties that interact with a Cloud. Business consumers benefit because they can consume and pay for only the resources that they need when they need them. 3rd Party vendors have a well-known reliable place to make their services available to Business consumers. Both business consumers and third-party vendors are subsequently able to offload a number of complex and expensive responsibilities with regards to the infrastructure that they would otherwise need to buy/lease, own, and support. Lastly, Cloud providers are able to provide the services that they have expert knowledge in and are able to receive massive bulk purchase pricing on physical infrastructure as they are required to house and operate so much of it to support their Cloud customers; everyone wins!

Returning to our conversation, the purpose of resource abstraction is to present higher-level applications services (the as-a-Service gang) with a set of what appear to be physical resources aka CPUs and memory, but in actuality are really just carefully allocated portions of the Cloud’s total very large resource pool(s). Of course, like all Cloud services, resources too are acquired as metered pay as you go items. Along with the computing resources that a consumer can acquire from a Cloud, they can also purchase all the storage (disk space) that is required for a consumers desired use. Lastly, all of the computing and resource services are rolled up and delivered to the client using virtualization. That is, a Virtual Machine that appears in all ways to both application and user as a remote instance of a computer. Humans can access these virtual machines via a Remote Desktop (RD) session, login and it appears to be the same as any physical computer that a user might log into remotely. But it’s not; “Magic”!!!

In the end, like most things in our sales and marketing slogan driven world, all of the grand items discussed previously herein actually just boil down to the same types of items that you might see in any private data center or group of data centers providing redundancy and close physical access to customers. Blade Servers (a set of computer boards typically housed in a rack enclosure) of some sort will be present in substantial quantity. These computers will all be connected to some sort of Cloud Storage Gateway (CSG) a software or hardware networking device that provides a standard method for accessing files, volumes and tapes that are housed in physical devices; the largest type of device being the Storage Area Networks (SAN). A given Cloud will have one or more SANs providing accesses to numerous Storage Array Enclosures (A hardware device that contains large numbers of disks) that contain many disks of different types including Hard Disk Drive (HDD), Solid State Drive (SSD) and even Self-Encrypting Drives (SED).

The final item left to cover in our tour of the Cloud is the network that all the virtual services we have discussed previously are made available over. At the start of our conversation I mentioned that a Cloud is effectively a “virtual” data center. To achieve this, each Cloud customer begins their Cloud implementation/presence by creating a Virtual Private Cloud (VPC). A VPC is a private secured network located inside a public Cloud that enables you to make use of the Public clouds metered services. Think of this virtual private network as being akin to a network that your business might have at any given data center only it’s actually residing in the Cloud.

Once the PVC is created, a network engineer can create the appropriate subnets (a subnet uses IP addressing to divide a network into a set of smaller networks that can be secured with and or from each other) needed for the services being placed into the Cloud. Then the engineer will secure the services via Access Control Lists (ACL) These lists specify what service(s) a given device can access both locally and on other networks or subnets. Lastly, the network engineer can configure virtual load balancers to allow network administrators to balance the traffic in your PVC properly based on the metered resources being used. And that is that, the Cloud as it stands in 2018.

There are numerous new Cloud trends evolving today so having a solid understanding of the base Cloud offering is essential to being properly equipped to take advantage of them. The Cloud of 2018 is a powerful group of technologies combined to help businesses to increase performance through the smart application of Cloud service offerings. I hope that you have found this overview helpful and that it makes your next conversation with your CTO, CIO, VP of technology or engineer just a little bit easier.