Digital IT Transformation and Cloud Computing

A digital transformation often involves a cloud but what does "cloud" really refer to? 

In addition to my role at Dell, I have been teaching graduate-level IT courses for 10 years, including cloud computing. Most of my students are existing IT professions, some very seasoned, and they span many different industries. Their companies may use various services from public cloud providers like AWS, Azure, or Google, but they sometimes have different interpretations of what makes up their in-house cloud.

To level-set everyone, both in our Dell cloud courses and my university courses, we always start with various cloud terms as defined by the National Institute of Standards and Technology (NIST - https://www.nist.gov). NIST defines cloud service and deployment models but my post today is about the essential cloud characteristics that are specified. They include the following:

A cloud could certainly have more characteristics, but these five are considered the most important.

When applying these to a public cloud, it is easy to understand since they have large self-service catalogs with many different offerings. The services are easy to access, and they have large pools of resources, including compute, network, storage, plus many more. Services can be setup to allow for scaling plus, they have features to view the services being used and the costs generated. Public clouds give the impression of unlimited resources and many of these characteristics are taken care of behind the scenes so the user doesn't have to think much about them.

In my opinion, I believe a good long-term goal for a private cloud is to behave like a public cloud provider but that is a pretty massive task. Where do they begin? Lets start with how the essential characteristics even apply to a in-house cloud, but we will save that for my next post...

-Bob

#IWork4Dell

As a follow-up to my last post where I mentioned about the five NIST cloud characteristics relative to a public cloud, we will now cover a little bit about these characteristics relative to a private cloud.

As a reminder: the NIST characteristics are: On-Demand Self-Service, Resource Pooling, Rapid Elasticity, Broad Network Access, and Measured Service.

In a private cloud, the good news is that Broad Network Access just refers to the people in your organization having access rather than being available to the full internet. Credentials still limit access internally as they do now.

Before we continue, I would like you to think about some of the IT capabilities available in your organization now.

• Are there virtualization services?
  • If so, there are likely virtual machines (VMs) that can be provisioned from a pool of compute resources.
  • Are there any virtualization capabilities for storage or networking?

• If someone needs a VM, how do they get it?
  • Do they have to send an email, or is there a service catalog they can go to and submit a request?
  • If there is a catalog, do you know if it just generates an email to IT and someone manually addresses the request or is the process fully automated? (You may not know, but I wanted to give you a few things to think about)

Resource Pooling can be done for compute, storage, network, databases, and so much more. Pooling enables equipment to be better utilized rather than the traditional way of different groups owning separate equipment that may only get utilized in bursts, or may go unused for months. Your organization may already be using resource pooling techniques even if they don't have a cloud.

A catalog may exist but is everything on the back-end fully automated to make it a true Self-Service Catalog? This requires orchestration tools and often the use of various application programming interface (APIs) and some upfront design work to make sure it all works as expected.

The best way to think about it is this: for the example of provisioning a VM, consider all of the steps it takes from the initial request to when the requester receives the provisioned VM. If it is a manual process, IT administrators might have their own steps or at least different order of steps that they perform. During a request, what if a new compute server needs to be added to accommodate the request. Rather than having the IT admin to do it, additional standby servers could be available and all of the steps to bring the server online, add it to the pool, and then provision a VM from it, could be automatically done through scripts or programming. By deciding on the best steps in advance and automating them, a more consistent experience is provided for the learner.

Rapid Scaling is interesting. Let's assume a company has a website and they have some equipment onsite to support the website (it could be the full website or just backend database servers to handle the orders).

• What happens if the traffic to the site and orders increase by 50%? Does the site crash and then the IT staff scrambles to add more systems to support the growth? If so, customers are lost.
• Ideally, the infrastructure should be in-place so that if there is a spike, systems can automatically be added to address the growth and prevent a crash.
• This takes planning but, in a public cloud, you can set it up where servers can be added or subtracted as demand increases or decreases so why not in a private cloud? It is certainly doable.

Measured Service is a more challenging one because although organizations often have some tools in place to log and measure pieces of their environment, they are often not complete when you look at them from the perspective of creating cloud services. If you have looked at or used any of the public cloud providers, you know that it it very easy determine what resources you are using, how they are being utilized, and any costs that you are incurring.

• How would you do the same for your environment?
• What tools or software would you need to add?
• How would you present this information to the user?

As organizations implement or enhance their private clouds and move toward providing their own cloud services, it is helpful to use these 5 NIST characteristics as a guide. Hopefully you are getting the sense that 'planning' is essential.   It is also helpful to investigate how the same features or services are offered by the public cloud providers.

-Bob

#IWork4Dell

Since my previous posts included details about National Institute of Standards and Technology (NIST - https://www.nist.gov), let's continue with the NIST definitions for cloud deployment models which defines four: public cloud, private cloud, community cloud, and hybrid cloud.

A public cloud is the easiest one to talk about; the best examples are Amazon Web Services (AWS), Microsoft Azure, and Google Cloud, but there are more. These providers have data centers across the globe and companies or organizations can use their services. Different companies can be using the provider's resources, and this is referred to multi-tenancy. If you have never explored any of the public clouds, I suggest you go to their websites to read about the many services that are available. Whether you want to quickly spin up a virtual machine, storage, database, or many other services, there are plenty to choose from. Some services are free but most services are available for a cost; the good part is you get charged for what you use and, when you don't need them anymore, you give them back (think of it like renting the resources). It is convenient because the provider owns and is responsible for all of the equipment.

Private cloud is for a single organization (I.E. a company); it is typically on the organization's premises but it could also be hosted offsite. If you think about an on-premise private cloud, the organization is responsible for all of the equipment and it's management, just like they do for their data center.

A community cloud is setup for a specific community or purpose and might be owned or managed by multiple groups in that community. Community clouds exists but are not that common. In my university course, we don't go beyond just the definition since most companies are interested in private, public or hybrid cloud environments.

A hybrid cloud is a combination of two or more separate clouds. Usually a hybrid cloud will consist of a private cloud and a public cloud, but it could include multiple private or even community clouds but that is less common.

A few years ago the term multi-cloud started being used but, in my opinion, multi-cloud is a type of hybrid cloud. Multi-cloud environments often have a private cloud plus two, or more, public clouds. Why would a company want to have multi-cloud? In the past, companies that decided to utilize services of a public provider first had to carefully plan to determine which provider made sense for them long term. Once selected, the company often became dependent on how that cloud vendor did things, this was referred to as 'vendor lock-in'. As all providers added more and more services, companies started to realize one vendor might be ideal or cheaper for some services (like backup or business tools) but other vendors might be better for different services (like big data, analytics, disaster recovery, etc.), thus the emergence of multi-cloud. Trying to connect any clouds, whether hybrid or multi-cloud certainly can be complex but, organizations are finding that there can be many benefits to doing so.

What I have found in the last 10+ years is that people sometimes think of 'cloud' only in the sense of 'public cloud'. If they know their organization is considering the 'cloud', they assume many things in their data center will be moved to AWS, Azure, Google, or another public provider and their skills won't be needed as much. However, if your organization is going to have an in-house private cloud (whether on it's own or hybrid, or multi-cloud), all of the compute, network, storage, and other equipment, plus databases, applications and most everything else needed will have to be setup, hosted, and managed. Many of the skills that IT professionals have today for administering their data center environments, can apply in a private cloud. There might be different tools, processes, and plenty of new things to learn, but there is strong overlap.

One benefit of a cloud is automation and orchestration which is sometimes a new concept for organizations. The cloud, and digital transformation in general, can also require a change in perspective; just because your IT organization has done things a certain way for so long, doesn't mean that is the best way to do them moving forward. Imagine what your IT team could do if many of the manual things that are done now could be automated.

If you look back on my previous post regarding the 5 cloud characteristics and think about your organization, you should be able to see where you might need to add to your personal skills.

Cloud is here to stay and it is an exciting area so explore and learn!

-Bob

#IWork4Dell

I wanted to go into a little more detail about multi-cloud from my previous post.

As I mentioned, multi-cloud environments typically consist of a private cloud and two or more public clouds. This allows organizations can take advantage of the services that they desire from each public provider whether the reason is for specific features, better performance, cost savings, etc.

If you have looked at any of the public providers, you know they have many offerings and features so it is understandable why companies have migrated to using services from different vendors rather than being tied to one (I.E. vendor lock-in). The different vendors all have their own way of doing thigs, such as how you connect to them, their user interface, command line interface, plus the programming interface used to access features through software (I.E. APIs).

How do you manage multi-cloud environments? One common way is that the different clouds are managed using their own tools, interfaces, etc. This means that an organization's IT staff not only needs to know the tools to manage their own private cloud but they also need to learn the tools available from any of the public providers they are utilizing.

Another way, and an ideal goal for managing a multi-cloud environment is to have one interface that would have visibility into all of the clouds. This means the interface would abstract the complexity and differences between the various clouds and make them appear more like one big cloud environment. Imagine having a workload running on your private cloud and you decide you want to move it to one of the public clouds, or you want something local backed up to one of the public clouds or to pull something from a backup, or you want to use some public analytical tools on some private data, etc. If you could do this through one interface makes it more of a true "multi-cloud" rather than "multiple clouds". Offerings exist to do this, including from Dell and VMware, but this is an area that will continue to grow especially because it make sense and will allow organizations to be more efficient and to best utilize their cloud environment.

-Bob

#IWork4Dell

What Shadow IT?                                                                                                 

According to Gartner, "Shadow IT refers to IT devices, software and services outside the ownership or control of IT organizations" https://www.gartner.com/en/information-technology/glossary/shadow 

The first time I heard the term in the context of a cloud environment was a little over ten years ago. This was when companies were discovering that some of their employees were using public cloud provider services (mostly AWS at that time) without their IT departments being involved. 

Why would employees do this back then and continue to do it today?  IT departments often have rigid rules and process in place and their staff may have more work than they can handle.  Therefore, when an individual, team, or business unit asks IT about the timeline for provisioning compute, network, storage, databases, etc. services, or purchasing new equipment, etc., the answer may be well beyond the timeline that they need.  If a team needs to produce a new application, product or service, the time-to-market may be essential to get a competitive advantage.  Since it is easy to 'rent' services from a cloud provider, the team may decide to forget IT and just get what they need from the provider.

This may seem logical, but it can be very risky.  When you provision services from your own data center, your IT department typically is handling the security, compliance, and any other policies or procedures that generally fall into the umbrella that is referred to as "governance".  You can use the resources that you need, even place company data on resources in your data center without much concern.

The same is not true for using resources from public cloud providers.  YOU are responsible for ensuring everything is protected and that is a lot of responsibility.

I recall the first year that I was teaching my university cloud course in 2012.  I had an AWS account and I created sub-accounts for all my students so they could start getting some hands-on experience to prepare for their projects.  Students started experimenting with VMs, storage, databases, and other resources.  Since they were all under my account, I received a email from AWS that several of the resources were not secure and were accessed by IP addresses from multiple countries; they wanted to let me know in-case that wasn't the intent.   Luckily, all the resources were just experimental but imagine if the same thing happened to the team above, especially if they had placed company data on the resources?

Another concern about using public cloud providers is that you have to pay attention to monthly costs because everything is 'pay-as-you-go', plus who is going to pay?  There are plenty of companies whose financial departments received bills passed on from teams using Shadow IT.  What if these costs were not part of the budget?

Many IT departments used to push back from having to do anything with the public cloud providers but once interest in hybrid and now multi-cloud environments grew, that changed.  Many IT departments realize the value from these providers and either have already or are in the process of figuring out how to connect and work with the public cloud services.

If you work at a company, and you are thinking about using any of the public cloud providers for any of your projects, I strongly recommend you reach out to your IT organization to find out their recommended best practices.

-Bob

#IWork4Dell

In our cloud certification courses we provide some common cloud related roles and their responsibilities.  Even in my university courses, where most students are already IT professionals, they like to know what is expected in various roles since many of their companies have digital transformation and cloud initiatives already underway.

The role names we use and who is responsible for what may vary in different organizations, but the responsibilities will provide insight into the types of skills needed.

In the next few weeks, I will highlight some of the more common cloud and IT transformation roles and explain their alignment to strategy, design, configuration, monitoring, and application development, etc.

However, before we do, the main thing to realize is that for an overall IT transformation and cloud project to be successful, teams must change from being independent silos to more cross-functional and collaborative.  To do this, individuals need to build knowledge is multiple areas as shown in the diagram.  There are plenty of domains besides what is shown but I will provide more information when I post each role.

-Bob

#IWork4Dell

The first role that I want to highlight is a cloud administrator.

Before we go into detail, think about system administrators that you know or have interacted with. Some might be knowledgeable in specific areas like: compute, network, storage, databases, web development, tools, or with specific technologies. Others might handle ordering or repairs of computer equipment, or setup web conferencing equipment. Some might be well versed with Windows or Linux, or work with critical applications the organization depends on; the list can go on an on.

For cloud admin's, the scope may be equally as broad and the specific responsibilities might be different responsibilities depending on the organization.

Cloud admins might be responsible for infrastructure, similar to system admins, but for cloud technologies. For compute, it might not be just hypervisors and VMs but also containers (like Docker) and container management systems (like Kubernetes). For storage, besides understanding various storage appliances, they may work with software defined storage or Storage as a service. For networking, besides working with physical network switches, hubs, etc., they might work with software defined networking. Instead of older style applications, they may work with cloud-native applications, or provide the tools and support that development teams need to build these applications.

Cloud environments benefit for automation and orchestration so some admins will focus on becoming experts in those areas or, since the goal is to offer services through a service catalog, some admins may be experts with catalogs.

Some cloud admins may be very knowledge with public cloud vendors like AWS, Azure, Google, and how their organization's private cloud can connect with these providers.

If you look at any of the main job posting sites and search for cloud admin, you will a wide range of skills listed.

What does this mean for you? It can mean a lot of opportunity. The areas I listed above are only a partial list of possible focus areas for a cloud admin.

Someone doesn't become a skilled system admin or cloud admin overnight, or by taking one class or training sessions, but learning more about any of the above areas is a good starting point.

In my university classes, most students are existing IT professionals, including some system admins. After learning more about the cloud with hands-on work, they realize a lot of the skills they have now, and even the way they debug and solve problems, can apply in a cloud environment.

--Bob

#IWork4Dell

Thank you so much for your posts and support @bobb! "Cloud = BobB"

The next role that I want to highlight is a cloud network engineer.

For some organizations, the responsibilities highlighted below might be assigned to cloud administrators who have network expertise. However, for this post, we will refer to the role as a cloud network engineer.

Responsibilities typically include:

• Maintaining and upgrading the network infrastructure on which the cloud operates
• Monitoring the performance and utilization of the cloud network infrastructure and services
• Ensuring the safety of the cloud network through various security standards and compliance.
• Integrating with the networks of other clouds to support hybrid or multi-cloud.

How are these responsibilities different from a typical network engineer?

• Hardware is still essential (I.E. network switches, routers, etc.), but more private cloud implementations are moving towards using software-defined capabilities, including software defined networking (SDN).
  • Software-defined is essential to automation which we will cover in a different job post
• In a hybrid or multi-cloud environment, a cloud network engineer will have to learn monitoring and performance services available from different public cloud providers as well as understand the underlying network capabilities to know how to connect the public clouds with their private cloud.

If you have an interest in networking or you are already a network professional, learning fundamentals about the cloud are essential since a recent survey shows that 76% of organizations are already using multi-cloud.

--Bob

#IWork4Dell

The next role that I want to highlight is a cloud engineer.

As I mentioned in my previous posts, the responsibilities for each role may be a little different depending on the organization or co mpany.

Think of a cloud engineer as a technical person that often does design and implementation work.

Responsibilities may include:

• Deploy cloud initiatives and educate various teams on how to implement these initiatives.
• Manage the engineering components of the cloud infrastructure.
• Orchestrate and automate cloud-based platforms.
• Design, develop, and debug cloud-based applications.
• Develop application programming interfaces (APIs).
• Connect cloud providers to the company’s network.
• Manage cloud environments based on the organization’s security guidelines.
• Stay updated on the latest trends and make recommendations.
• Formulate best practices for the cloud team and the entire organization.
• Work with engineering teams to determine the best-suited cloud solutions for the organization
• Create and maintain a disaster recovery plan.

As you can tell, this is a very big list so some organizations may have engineers that focus on specific areas and might be called something else.  As I post more roles, you will see that some of these responsibilities overlap.

If you think you might like to be a cloud engineers, there are certainly plenty of areas to focus on!

--Bob

#IWork4Dell

The next role that I want to highlight is a Cloud Application Developer

(As previously noted, the responsibilities for each role may be a little different depending on the organization.)

Responsibilities typically include:

• Integrate the various cloud services through application programming interfaces (APIs).
• Develop custom applications within a cloud environment using the organizations lifecycle.
• Design and develop secure cloud products, services, and applications that include:
  • Front/back end, full stack, and web applications.
• Enable cloud application deployment.
• Support, monitor, maintain, and improve the applications.

The focus should be on building cloud-native applications rather than the traditional monolithic ones.

Also, since the goal with cloud is self-service, the applications ideally should incorporate as much automation as possible and even development environments can be automated.

As expected, a software background is very helpful! Examples of applications include: mobile (think of your phone’s app store), gaming, SaaS offerings, internal to your organization (I.E. business applications, analytics), even apps just for the IT department to help manage the private cloud, plus many more...

FYI: The topics in green will be covered in future posts

--Bob

#IWork4Dell

"What is a Service Lifecycle"

I decided to mix it up a little in this post.

In my previous post about the role of a Cloud Application Developer, I mentioned that one of the responsibilities may include "Develop custom applications within a cloud environment using the organization's lifecycle".

What is a lifecycle? Some of our courses contain examples of cloud service lifecycles, including this one:

We use it to help learners understand the process of going from thinking about, to planning, to building, and then running cloud services, but, if you think about it, many of the stages in this lifecycle apply to so many things.

Lets start with something non-technical: This lifecycle would 'mostly work' if you want to build a small shed or greenhouse in your yard. You would certainly start with planning and go through the design, and build. Test, launch and operate would be simple (maybe just put stuff in the shed or start growing flowers) and the feedback might be limited to your neighbors either happy that you have a place to hide your stuff or they are not happy because they have to look at the new structure.

Back to technical: Assume you are a mobile app developer and you come up with a great idea for an app for your company. If you are going to get support so you can have the time and money to build it, you have to come up with a strategy or plan to show that it makes sense. Once you get the support, you have to do a more detailed design, then build, and test as you make progress. When it is getting closer to being ready, you may launch it to a small test group and to test on many mobile devices, then make changes, but eventually you will launch it for real. When it is running, you want to make sure it is running as expected and you want feedback so you can continue the cycle to fix and update it.

I haven't provided enough information yet to start talking in detail about cloud services (so many future topics!). However, since this is a general discussion cloud thread, it might be fun to start thinking about how many times a day that you interact with something 'cloud related'. It could be just checking your Gmail, or using Office 365, you might order something from Amazon or some other online retailer. Have you looked at any real estate and used the map function to see where the property was? If you watch sports, sometimes they post statistics or analytics and you see on the screen "powered by AWS". How many phone apps to do touch on a daily basis? Or websites do you visit? Have you recently taken a ride-share service? Did you have work done this week at your house, maybe by a plumber, HVAC technician, cable installer, electrician? Did they send you a bill using a cloud app or update your customer information via an app? Do you use Alexa, Google translate, or Siri? The list goes on an on!

As I post more job related job roles, hopefully this information provides a little more context and I will tie some of the roles back to the lifecycle.

-Bob

#IWork4Dell

The next role that I will cover is Cloud Architect.

Like with many professions, an architect is an important role because they are responsible for determining how all pieces will fit and work together. They are key decision makers and provide guidance.

In our cloud courses, we highlight some responsibilities for a cloud admin including some that are specified by Gartner.

The day-to-day responsibilities include:
• Lead the cultural change in an organization's cloud adoption.
• Plan a cloud strategy and coordinate adoption.
• Develop and maintain the cloud architecture, including hardware, software, applications, etc.
• Determine best practices for the organization's cloud.
• Estimate and manage cloud costs.
• Collaborate with IT security team to monitor privacy.
• Select cloud service provides and third-party services.
• Create and implement a multi-year road map for the cloud platform.

The public cloud vendors and a number of private vendors, including Dell, offer cloud certification offerings. They are a great way to get started but, like any lead role, it takes time. However, these programs can help if you decide you would like to pursue some type of a cloud career.

If you were to talk with 10 cloud architects, you will likely find their career journeys to be different. Some might have been system admins, others might have focused on compute, networking or storage, others maybe tools and automation, and I could list a bunch more. So, if a cloud architect intrigues you enough, look into it more; you might find it very interesting!

-Bob

#IWork4Dell

The next role that I want to highlight is a Cloud Security Manager.

This is the emerging role, and the responsibilities can vary from organization to
organization. Since the cloud service providers have their own security teams, a
cloud security manager usually coordinates with them to strategize and align
efforts. They must work out different strategies to fulfill compliance requirements
which may not align smoothly with the providers’ offerings

This role may involve the following:

• Responsible for the complete security and risk management strategies of a
  cloud environment.
  
• Designs and employs security measures for the cloud infrastructure, plus all the
  products, services, and applications running on the cloud.
  
• Ensures that security standards and compliance requirements are met.
• Works with cloud providers to plan and implement security policies within the
  organization.
  
• Understands the risks of security failures and their technical and
  business implications.
  
• Manages access to cloud resources using user and group accounts.

You can image this gets more complicated in multi-cloud environments!

--Bob
#IWork4Dell

Digitalization of business is very important because the more automated the business, the better