Hello and welcome to this introduction to cloud technology sales class in which I will be covering some of the key concepts of cloud computing. Cloud deployment models, various cloud service models. Advantages and disadvantages of cloud computing, and other related matters that will be useful to individuals who are looking to pursue a career in cloud technology sales. Cloud computing is the on-demand delivery of computing, storage, networking, and other IT resources via the Internet with a pay-as-you-go pricing. So it follows a utility, a gas or water type business model, where one can turn on computing resources and pay for the usage of those resources, and then turn them off when done. Before cloud computing, organizations and businesses and agencies had to procure a lot of hardware. Compute hardware, storage hardware, networking hardware and place them in controlled environments with power cooling, physical security staff. So those tremendous upfront capital expenditure. And other issues would come after applications were released when organizations had to guess as to how many clients, how many people would be hitting their websites or their ecommerce stores. And did they have enough capacity? Did they have enough storage? Now with cloud computing, one can think of the infrastructure not as hardware, but as software resources we can change very, very quickly. Computers can be turned on, storage systems can be brought online, and then turned off or taken down when the demand has subsided. So it's a much more flexible model of computing that has really alleviated a lot of issues that businesses and organizations used to have in the past. Everyone is exposed to cloud computing applications in various forms, whether they're social media apps like Facebook and Twitter. Storage, Dropbox, OneDrive, streaming platforms, communication platforms, productivity suites, Office 365, now Microsoft 365, Zoom. And other business applications such as Salesforce, Marketo, Hubspot. These are all examples of cloud computing. And the three main cloud infrastructure providers are Amazon Web Services or AWS, Google Cloud Platform, and Microsoft Azure. There were a lot of, Cloud-like computing available in the 60s and 70s through the 90s when people were able to use mainframes to compute and run analyses. And in the 90s, a lot of networking services were implemented by telco companies for efficiencies. These services with software-based. Around 2006 is when the birth of the modern cloud took place when Amazon released S3, the Simple Storage Service, and EC2, Elastic Compute Cloud. Soon followed by Google and Microsoft and other vendors around the world. But it was really when Amazon released the Amazon Web Services is considered the birth of modern cloud computing. The cloud deployment models can be broken to three different categories, where if an organization is using everything at one of the vendors such as AWS or Azure, it's considered a public cloud. And the organization essentially is treating the cloud vendor infrastructure as a data center, where they're renting services, compute, storage, and networking services at the vendor sites. And it could be IBM, Alibaba, or other kind of niche or very specific vendors that run their services through a cloud infrastructure. A private cloud is one in which organizations are restricted either by privacy concerns or data concerns, whereby they have to have everything in-house. And they can still use third party cloud services, but it's maintained on a private network. Thereby providing the right regulatory functions and privacy and security implementations that are required by the organization. Lately, there are a lot of hybrid cloud deployment models whereby organizations are using one or multiple public clouds along with private clouds and private data centers. And having secure ways of communication between these various deployment models. For each deployment models, there are various service models whereby organizations can interact with the cloud vendor as they choose. The three most common ones are infrastructure as a service or IaaS whereby a client essentially accesses or maintain servers and operating systems at the cloud vendor. And the cloud provider, the cloud vendor really, other than providing the bare metal and the networking and handling all the security and electric connections to these machines really don't access the machines at all. Platform as a service or PaaS is meant for primarily allowing the clients to build and run applications. There are various platform as a service providers out there, including Force.com, Heroku, Amazon Elastic Beanstalk. In these cases, the provider actually manages the patching and the setup of the underlying operating systems, database service, things like that. And the client just focus on building and running applications. The third service model is software as a service or SaaS, which is where the provider manages everything from the hardware, all the way to the entire product and the infrastructure. And the client just uses the software which is provided as a service and is only responsible for access to these applications and for clients' proprietary data. So things such as Microsoft 365, Salesforce, Slack, Zoom, Dropbox. These are all software as a service models that are being generated on the underlying cloud platform. As you go up the stack from IaaS to PaaS to SaaS, the levels of abstraction increase. And as you'll see later on are a lot harder to get off the higher up the stack you go. This chart is comparison table provided by BMC shows what a client would manage in blue versus what the cloud vendor would manage in orange. As you move along the stack from having on-premises applications and services to infrastructure as a service, where the vendor just manages the networking and the underlying hardware. And the client manages the operating system on up. Platform as a service, where the client manages the data and the application, they build the application. Software as a service, where the client just manages access to the application. In addition to these three main services, there are other services that are more niche, function as a service, Blockchain, mobile backend. And in particular, function as a service, it's now being subsumed under a broader term that goes by serverless computing. Which is a model that is used by the cloud provider to actually dynamically allocate the machine resources to the client. The client doesn't even have to worry about the underlying hardware. And all the client needs to do is upload their code and the serverless application will then handle what needs to be provisioned based on the code that is uploaded by the client. And it's becoming more and more prevalent with AWS Lambda and Azure Functions or serverless databases such as Amazon Aurora or Azure Data Lake. Things such as Amazon Alexa and other kind of speech recognition application that are out there are based upon these serverless services that take in some information. And then instantaneously or a neural time provide a response. Some advantages of cloud computing. In addition to kind of the flexibility in the agility that clients now have to access hundreds and thousands of servers at a moment's notice and have those servers have a global reach, which was never the case before. The client really does not have significant upfront cost. The cost is a variable ongoing cost and no longer a capital investment that needs to be made and planned for and waited on for the resources to be then shipped to the client's data centers. So a lot of the overhead in kind of the operational issues associated with managing a data center no longer exist in the day of cloud computing. And clients and organizations can focus on what they're good at and focus on innovation and focus on solving business problems rather than worrying about their data centers and, Global reach and predicting the capacity that was going to be required. Now, all of that can be done automatically and with software solution that allow for great flexibility. Some of the disadvantages of cloud computing. As I alluded to earlier, there's vendor lock-in and there's cost overrun. Vendor lock-in is always an issue with software. Any technology and hardware, any technology that a business procures, they're locked in to some level to using that. Including open source free software, you're still locked in. There is a time component associated with changing. But it's more so with cloud computing because they make it very easy to move all your information in, but there is a cost associated with taking it out. And the more sophisticated, the more up the stack you go, software as a service, the tighter the level of lock-in is. And it's much harder to get out of, say, Microsoft 365 and just to get out of a storage and compute environment like infrastructure as a service, where you just are managing all the service. You can go to another cloud vendor and procure the same service. Cost overrun is also a huge issue because it's so easy to turn things on and a lot of times people forget about it. And developers are not used to tracking spend like IT and other operational folks are. And so there is a lot of wasted spend associated with kind of the ease of use of cloud computing. But vendors have taken this information to heart and have started implementing ways in which organizations can turn on alerts and can scan systems to see what's being unused. And then turn them off automatically as well, thereby maintaining and managing their cloud spend. Some other topics that will be covered include security, very, very important concept because it's a shared responsibility model. It's a partnership between an organization and the cloud vendor. And the vendor is responsible for security of the cloud, the hardware, the HVAC system, the physical security. And depending on the solution and the service that's being provided, database patching, operating system patching, things like that. While the client or the user is responsible for security in the cloud, essentially access to the services that the client is using. And the data and related encryption of the data that is important for the client's requirements. We'll also talk about migration to the cloud and the different methods that are available from a standard lift and shift. Just mimicking what you have in the data center on site to the cloud, all the way through containerization and cloud native architectures, which really take advantage of what the cloud vendors do provide. We'll talk about engineering for the cloud, different engineering models, Agile, DevOps. As well as CI/CD, whereby you can do continuous integration and deployment with various cloud services that are provided for automated testing and pushing to live systems. Cloud expenses, ongoing expenses where there's total cost of ownership is a very, very important topic, both from a procurement standpoint, as well as when clients use these on an ongoing basis.
