An Intro to Cloud Computing
Before we start looking at the specifics of the different types of cloud adoption strategies, let’s touch upon the various shapes of cloud computing that are available for consumption today.
There are three major public cloud providers, also known as hyperscalers: Amazon’s AWS, Microsoft Azure and Google Cloud. The meteoric rise of these three providers over the last five years is a clear testament to the upward trend of cloud adoption.
The cloud is also delivered in a number or different computing models—including infrastructure as a service (IaaS), platform as a service (PaaS), and software as a service (SaaS). These
The cloud also can be leveraged using different deployment architectures—private cloud, public cloud, hybrid cloud, multicloud, or combinations of those options.
But before you consider which of these options is the best fit for your cloud adoption strategy, you should start by asking a simple question: “Why do I need to use the cloud?”
Start by Asking “Why the Cloud?”
The success of a cloud adoption can depend on understanding why your organization needs to move to the cloud.
Anyone looking to successfully adopt some form of cloud computing model should start with establishing the clear business requirements and the expected business outcomes first. Many unplanned cloud adoptions can risk failure due to costs spiraling out of control and other factors. A poorly planned migration can prove to be a significant risk and potentially cause irreversible damage to the credibility of the IT departments within the organization.
The best way to avoid such outcomes is to start with the “Why?” Why do you need to use the cloud?
Are the costs of on-premises deployment no longer scalable for your business?
Is your organization looking to reduce business risk with the improved infrastructure reliability and availability of the cloud?
Are you trying to grow revenue by quickly creating new products and services and offering them to new markets through a geographically distributed public cloud platform?
Is your organization looking to increase its operational efficiency and improved productivity by automated, self-serviceable cloud IT?
While every organization's answer will be different, it is important to clearly establish, document, and assess the business drivers that would underpin the adoption of cloud as the starting point to any form of cloud adoption plan and get the business leader’s buy-in.
Getting Answers from the Cloud
Fortunately, all the main public cloud providers provide cloud adoption frameworks that can help customers who are getting started on this journey.
For example, the first step of the AWS Cloud Adoption Framework (CAF) talks about the “Envision” stage as the beginning of the journey, where customers are encouraged to identify and prioritize the cloud transformation needs in line with specific business objectives.
Similarly, the Azure Cloud Adoption Framework advises customers to start with the “Strategy” stage where defining the clear motivations for cloud adoption and classifying them based on business needs. These business needs can include planning around a critical business event, such as, for example, a migration aimed at scaling to meet market demand or the release of a new service.
Many third-party systems integrators or solution providers also offer various assessments and consultancy services (usually at a cost) to help customers get started with establishing the “why” before embarking on a cloud adoption project.
It is important for customers who are considering such services to ensure that the system integrator or the solutions provider partner chosen to provide these services are competent and skilled across all the key cloud computing models as well as the deployment types in order to receive impartial advice and guidance. Learn more in Four Key Considerations for Choosing Your Cloud Managed Service Providers.
During these engagements, which may consist of multiple stages, the solutions provider would look to identify the business drivers through consulting the business leaders, followed by identifying potential opportunities, risks, and the cost of adoption across various cloud platforms and deployment models.
Building a sound business case based on such outputs and getting the key business stakeholders’ commitment to execute can help ensure a successful adoption, and most importantly, the actual realization of the intended business outcomes of the cloud adoption.
Let’s take a closer look at some of the popular ways to adopt some of the cloud computing models and the deployment types based on popularity and what are key points to bear in mind.
Choosing the Right Computing and Deployment Models for Your Cloud Adoption Strategy
Once you’ve answered why you need to adopt the cloud, it’s time to consider what type of adoption architecture would be the most beneficial to meet that goal.
Hybrid Cloud Adoption
IDC research shows that hybrid cloud deployment is an inevitability for most enterprises given their extensive use of the on-premises data centers as well as public cloud platforms.
Many enterprise customers tend to have years of technology investments in their on-premises data center based infrastructures. Most of these organizations and their internal culture are driven by their People, Product, and Processes. That means they have been collectively designed over the years to support the on-premises solutions approach.
Migrating the entirety of that on-premises technology stack to a cloud platform is a huge challenge. It means transforming the associated internal IT processes to support the operations, governance, and compliance of the cloud-based environment. Also, upskilling people within the organization to support that new environment takes significant amounts of time. In some instances, it can even be unrealistic to do so.
For such organizations, the hybrid cloud approach can be much more suitable. For these customers, hybrid cloud provides the most business benefits at the cost of minimal transformational risk.
The first step to hybrid cloud adoption typically involves identifying specific use cases that are suitable for transferring to the public cloud and producing an effective adoption strategy for each of these use cases. Fortunately, there are many tried and tested use cases that customers can begin with. Let's have a look at some examples:
Infrequently accessed data: Leveraging cost efficient, scalable cloud storage for storing infrequently accessed data (such as backups, archives, and file shares) has been a proven starting point for many organizations.
One of the biggest challenges faced by enterprise organizations with such data has been the ever-growing storage requirements and ongoing operational costs to host and support that data.
Cloud object storage solutions such as Amazon S3 Storage, Azure Blob, and Google Cloud Storage provide simple, infinitely scalable, fully managed storage solutions with various tiers of object storage that can store this infrequently used data at a much improved price point. In the case of backup data, object storage also satisfies the 3-2-1 strategy for backups.
Disaster Recovery: Similar to backups, disaster recovery is also mandated by various regulatory requirements for many businesses. Due to the increased dependency of most organizations on their digital IT platform, which can underpin every aspect of their business, having a credible disaster recovery solution is not just limited to large enterprises.
The single biggest challenge that these customers face with their disaster recovery solution is the underlying infrastructure costs of the secondary data center required for the disaster recovery. Maintaining a second data center to replicate data, with sufficient infrastructure resources to run all their mission-critical systems in the event of a disaster (that may not ever happen) is an expensive prospect for many.
Cloud computing provides a simple solution to this problem. Users can replicate their mission-critical data to a cloud platform with minimal compute requirements. These resources can be scaled up during an actual disaster event, significantly lowering the total cost of ownership of their disaster recovery strategy.
Development and Testing: Similar to the use cases above, cloud computing also provides a simple solution for transient requirements such as application development and testing.
Many organizations that develop applications internally require dedicated infrastructure platforms that are often short lived during the application development and testing processes. Having to maintain dedicated on-premises infrastructure that is isolated from the production environment requires significant duplicate infrastructure costs.
Leveraging cloud computing models such as IaaS enables organizations to provision various development and testing environments on demand, which can be scaled up and down based on specific development and testing needs, and help save costs.
Hybrid Deployment Challenges and Solutions
When it comes to planning a hybrid cloud adoption strategy for these (and many other similar) use cases, customers must be aware of some of the challenges they can encounter.
Getting insight into the data set
One such challenge is the complexity involved in the identification of such infrequently accessed data. With variable access patterns, identifying which data is no longer accessed (e.g., a set of documents or images from a specific historical project). That would require continuous scanning and analysis of multiple file shares in a geographically distributed enterprise organization, which is significantly challenging in itself.
Reducing migration complexity and costs
Once identified, migrating all of this data to the cloud (and back, if needed) within each application could also face technical complexities and cost constraints. Users must also consider the egress costs every cloud platform charge, which can make data movement between repositories for occasional access costly.
To solve these issues, users can make sure the migration process and data transfers are efficient, thereby reducing ingress and egress costs as well as the required connectivity, bandwidth, and transfer time. Similarly, such application specific data can be subject to platform locking which can limit the mobility of data across platforms or add additional complexity in the form of data conversions required.
Gaining visibility and control
In addition, once data has been migrated to a cloud platform, most customers are now faced with having to run and manage these different data estates across multiple infrastructure platforms using different management platforms and the associated toolsets. This can add various operational complexities and increase the probability of human error and various governance issues due to lack of holistic visibility to their overall data estate.
Customers considering hybrid cloud adoption need to have a clear strategy covering these challenges and many more. As discussed above, starting with the “Why” helps establish the clear business requirements for the cloud migration. Once that’s answered, it’s time to consider the “What” and the “How” of the migration.
The “What” needs to define the various tactical and strategic workloads that need and can be migrated to the cloud. Tactical workloads can include simple, tried and tested workloads such as DR, infrequently used data, etc. Strategic workloads include key application data, user data, etc. What might not move? Any data that are subject to specific restrictions such as data sovereignty or privacy regulations can be clearly identified and left running on-premises in proprietary data centers owned and operated by the organization.
The “How” of the cloud migration should define the specific data migration and the ongoing management and operation processes for the identified workloads to be moved to the cloud. It is key for the customers to follow the appropriate cloud adoption frameworks (mentioned above), factor in the challenges, and have a plan that mitigates those challenges for their adoption process to be a success.
Using Data Management Platforms to Enable Hybrid Deployment
Hybrid challenges, such as those discussed above, can be easily overcome by leveraging an overarching cloud data management solution. A data management platform can provide an organization with the visibility into its overall data estate, spanning both on-premises and the cloud platforms, while mitigating the risk of platform lock-in.
Data management platforms can also help by providing a single, storage-based data migration and replication technology that leverages built-in data efficiencies such as deduplication, compression, and incremental forever data replication. This can reduce migration complexity and costs.
Having an effective hybrid cloud strategy incorporating these kinds of mitigative capabilities can ensure enterprise customers realize better value of their hybrid cloud business plans.
Similar to the hybrid cloud, multicloud adoption has increasingly become popular over the last few years, especially at the enterprise level.
Multicloud adoption makes it possible to utilize specific, best-of-breed technologies from each cloud platform while diversifying the risks of being tied down to one cloud platform. These benefits have driven the adoption of the multicloud model, even despite some of the operational challenges it can pose.
The challenges that come with multicloud adoption can include:
Each cloud platform leveraging proprietary technologies for processing and storing customer data, which restricts mobility of the same data across platforms and various egress costs can even add a transfer penalty.
Users are forced to work with multiple cloud portals for day-to-day management which can add to the operational complexity of the IT teams.
Developers working with each cloud platform will also be influenced by proprietary application and compute technologies (e.g., AWS Lambda versus Azure Functions) which can create deviations within various development teams in the same organization and can lead to technical and operational compatibility issues.
In order to avoid or mitigate these challenges, enterprise customers considering the adoption of multiple cloud platforms can focus on having a common application framework and a common data management layer that are compatible with every cloud platform.
A common data layer that can cater for the storage needs of various different applications, across all cloud platforms that is managed through a single pane of glass can simplify the operational complexities. It can also help reduce the platform locking issues by relying less on the underlying cloud storage platforms and can facilitate easier data mobility between the clouds which can benefit these customers.
Similarly, using a common application layer can provide a uniform application stack across all the cloud platforms that would also benefit multicloud customers. Such is the case with Kubernetes.
Containerized applications based on Kubernetes are increasing in popularity and every major cloud platform now has a fully managed Kubernetes platform. Additional details can be found about Amazon Elastic Kubernetes Service (Amazon EKS), Azure Kubernetes Service (AKS), and Google Kubernetes Engine (GKE) where customers can deploy and run their Kubernetes workloads easily, breaking barriers to entry for many.
Any Kubernetes based workload can be easily deployed on any of these Kubernetes environments which provides customers the critical application portability to deploy their business-critical applications across multiple cloud platforms and facilitate migrations without the need to develop applications specifically for each cloud platform.
Google Anthos, Azure Arc, and Amazon EKS and ECS Anywhere provide ways for users to overcome the limitations of hybrid and multicloud deployments. These services can be used to deploy resources within the respective cloud vendor ecosystem, and also extend the use of those resources to on-premises and other cloud vendor environments. While EKS, AKS, and GKE offer some level of portability, services such as Anthos, Azure Arc, and Anywhere, enable a full hybrid and multicloud experience.
Cloud-Native, Cloud-First, and Cloud-Only Adoption
Many customers looking at the cloud are starting to produce various strategies prioritizing the use of cloud over on-premises. Cloud first, cloud only, and cloud native are popular varieties of this approach. Let's take a closer look at each.
Cloud Native Adoption
Cloud native typically refers to new ways of application development powered by cloud technologies and cloud principles along with the associated processes optimized for the automation and the agility provided inherently via the cloud computing model.
While it’s true that cloud computing models such as PaaS, SaaS, and even IaaS can be broadly classified as cloud native, the generally used context for “cloud native” revolves around building loosely coupled applications that consist of a collection of small services, aka microservices. These microservices are distributed, resilient, manageable, and observable with a high degree of automation for continuous integration and continuous delivery (CI/CD) pipelines as an application development method as defined by the Cloud Native Computing Foundation.
Containerization plays a key part as a must-have prerequisite technology required for cloud-native applications and container orchestration technologies such as Kubernetes also play an important role in making cloud-native applications viable at scale.
Given the micro and mostly stateless architecture of these applications—which can be spun up and down based on demand in a fully automated fashion—they are ideally suited for the cloud computing cost model.
However, customers considering cloud-native adoption need to be mindful of the significant application transformation (re-development) costs that can be associated with migrating their existing monolithic applications into a microservices architecture. Due to the technical complexities involved, most customers prefer to adopt the cloud-native approach for new application development requirements, which can be architected from the ground up with this model in mind and are ideally suited to leverage the cloud.
Cloud-native applications are also increasingly being used for application portability purposes across multiple clouds. However, users can face a challenge when stateful applications on one cloud platform require access to external data sources (such as Amazon S3 for EKS applications) and may not function properly once ported over to another cloud platform. Leveraging a data management platform can help by making such persistent data (e.g., a database file or a file share) easily replicated and available across all the cloud platforms in use.
A cloud-first strategy is one where the use of cloud-based solutions is preferred as the priority over any other form, such as on-premises based solutions. Cloud-first strategy promotes building applications or business solutions in the cloud to begin with, however it does not restrict use only to the cloud.
Many organizations have started to mandate the cloud-first approach as a corporate strategy for all their future IT needs, and many governments around the world (such as in the United Kingdom) have also mandated cloud-first adoption as a policy for all their public sector arms.
Customers considering cloud-first adoption strategies need to examine the benefits, but also the potential challenges. While any newly formed organization must consider the cloud its first and only choice for deploying applications, some may be prevented from doing so due to security, privacy, and data sovereignty requirements that would mandate customer-owned and self-managed IT infrastructures.
In addition to such policy-driven challenges, existing enterprise customers embarking on cloud-first adoption need to consider the requirements for their legacy systems and legacy IT infrastructure on-premises to co-exist alongside their cloud infrastructure. This can be challenging due to juggling management platforms, lack of overall visibility, and the lack of appropriate cloud management tools to support multiple platforms.
Unlike the cloud-first approach, cloud-only adoption enforces the cloud (often public cloud) as the sole option for all IT requirements. Cloud-based IaaS, SaaS, and PaaS solutions play a vital role in enabling customers to embrace cloud only strategies due to the depth and breadth of the solutions available.
Many customers have successfully moved away from the traditional, self-managed enterprise applications such as Microsoft Exchange in favor of the cloud-based SaaS platforms, like Microsoft 365. The meteoric rise of Salesforce as a popular SaaS-based CRM platform is another example of customers starting to move increasingly to the cloud-only approach. Similarly, the growth of IaaS cloud solutions, such as AWS EC2, Azure Virtual Machines, and Google Cloud Compute Engine over the last few years indicate the increasing popularity of the cloud-only option.
While a cloud only strategy may be a good fit in some instances, customers should be aware that cloud-only strategies can sometimes be limiting and may not provide the full flexibility provided by an alternative strategy, such as the hybrid cloud approach. Enterprise customers in particular may find it difficult to embrace the cloud-only strategy due to issues with integration, service availability, data migration, and cyber security constraints.
Hybrid Cloud Vs. Cloud-Native/Cloud-First/Cloud-Only
Hybrid cloud deployment typically refers to an adoption model whereby customers are happy to continue to utilize both on-premises as well as public cloud resources. This allows them to benefit from the best of both platforms.
While cloud-native, cloud-first and cloud-only strategies can also include hybrid cloud type deployments—especially during the initial transient stage of migration when both platforms will be in use—those strategies all intend to permanently move away from the on-premises IT model. It is therefore very important for customers to carefully consider the pros and cons of each strategy discussed above, in order to clearly establish the viability of that final aspiration of moving all of the on-premises application and the infrastructure stack to the public cloud.
The final word on the difference between the hybrid approach and the other deployment models is choice. The hybrid approach allows you to decide on a case-by-case basis (i.e., per app, per workload, etc.) which environment is best fit given practical considerations. Cloud-native/first/only can be considered much more restrictive and less flexible in that regard.
Migrating to the Cloud
Customers considering any of the cloud adoption models mentioned above need to consider the pros and cons of each approach.
While the cloud-only or cloud-native adoption models may be perfectly suitable for a startup with no traditional IT platforms to maintain, an existing enterprise customer with many monolithic applications with years of technology investment may not fully realize any business benefits by migrating those workloads to the cloud as is.
In most instances, these customers will be faced with significant application transformation costs when re-architecting these applications to be cloud native. And due to technical complexities, such transformation costs could significantly outweigh the benefits to be gained from running them in the cloud.
Enterprises looking to move to the cloud should always consider the hybrid cloud adoption model in order to continue to support their traditional applications on-premises where they might be better suited. However, those data centers can now benefit from the efficiencies introduced by leveraging cloud-based technologies. These could range from key use cases, such as migrating or tiering infrequently accessed or even Tier 2 data from the on-premises data centers to the cloud as well as leveraging the cloud for disaster recovery purposes realizing significant operational cost savings.
In addition, these users can also leverage new cloud capabilities such as data analytics through migrating key data sets to the cloud without the need to build complex analytics platforms on-premises. These organizations can also consider cloud for any new solution requirements and cloud native technologies can be an ideal solution for these new applications.
There are a number of different cloud computing models and deployment types available for organizations to consider. These computing models and deployment types can be adopted in many different ways, and each has pros and cons that need to be considered beforehand.
While each major cloud platform provider has a specific cloud adoption framework with specific stages to help customers get started, there are also other third party, cloud agnostic frameworks available from various solutions providers and systems integrators to consider. Generally, all frameworks advise the customers to begin with understanding the business requirements and defining a clear strategy that includes the most suitable adoption model, along with a clear plan and the business milestones to evaluate the successful adoption.
Out of all of these options, there is one choice that is most feasible for enterprise organizations, especially those with existing on-premises technology solutions: the hybrid cloud approach. Hybrid cloud architectures allow enterprise users to benefit from the best of breed technologies from both cloud and on-premises and continue to meet compliance requirements around security, privacy, and data sovereignty.
When embracing hybrid cloud for these enterprises, there are common challenges such as complicated data migration, platform locking, and lack of centralized management.
Cloud Volumes ONTAP spans across the entire spectrum of the public cloud service providers, with access to AWS, Azure, and Google Cloud, while seamlessly integrating with existing ONTAP deployments on-prem. This makes it easy for Cloud Volumes ONTAP to orchestrate hybrid and multicloud deployment models and help you meet their challenges head on.
If you’re looking to accelerate your cloud journey and step into the cloud with a hybrid deployment, consider signing up for a free trial of Cloud Volumes ONTAP or talk to one of our cloud solutions architects today.
What are the major reasons for cloud adoption?
There are many reasons for adopting the cloud, but the most significant factors are those related to reducing costs, keeping up with the latest technology, and working more easily across disparate locations.
What are the cloud adoption phases?
No two migrations are the same since every cloud adoption strategy will differ depending on the use case. However, there are some broadly defined phases that generally every adoption will go through.
The first phase, in the most general sense, is to understand the business need for moving to the cloud—the “why” of the migration. Next, the “what” has to be answered: determining which parts of the existing on-prem deployment need to move to the cloud. After that, “how” the migration will take place needs to be decided.
How do you promote cloud adoption?
Organizations trying to promote cloud adoption need to identify the principal stakeholders who will be affected by and can help influence and carry out the adoption. These “cloud evangelists'' can help advance the project goals, win over holdouts, and can come from departments as disparate as IT and the financial departments. Sometimes, the decision to move to the cloud comes from top management.