Hybrid and Multi-Cloud Perspectives: 2020 trends driving deliberate, strategic adoption
Whether by implementation of strategic technology imperatives or by accident, multi-cloud is here to stay in the technology portfolios of the world’s largest IT organizations. IT is quickly moving toward the Data and Digital Platform (DDP) of the future to establish the coherent applications, data and technologies required to answer business needs and enable the delivery of business use cases rapidly and at scale (1). Once established, a DDP allows on a defined perimeter to assemble ready to use, reusable components and hosting common capabilities, as well as offer standard interfaces to integrate new internal and external services easily. One of the most foundational components required to establish DDP includes cloud-native infrastructure in a multi-cloud deployment model, where all core applications, data environments, and digital applications are built and reside upon.
Figure 1: Cloud is a key component that underpins the Next Generation Architecture of the Data and Digital Platform (DDP)
The demands of platform agnostic digital products, workload portability, encapsulation of data via API-led connectivity, modularity, and a decoupled technology stack have all driven many IT organizations towards a multi-cloud posture. Understanding the drivers behind this trend, why true multi-cloud is difficult to enable, and how to enable the portfolio for this new delivery model is critical to maintain an edge on the most flexible, market leading operating model in the cloud space in 2020.
2 What is hybrid vs. multi-cloud, and why is it an increasing trend?
2.1 What is hybrid and multi-cloud?
First, let’s explore what multi-cloud is as a cloud deployment model and how it differs from other cloud environment options in order to better explore and understand the tradeoffs for making architectural choices.
Hybrid cloud is a deployment model where DDP and business applications are built on both on-premises infrastructure in a client Data Center, and a single public cloud environment (e.g. Amazon Web Services, Microsoft Azure, Google Cloud Platform). All of the major public cloud service providers (CSP) have established hybrid cloud offerings in order to extend and integrate certain environment services and standards, to include AWS Outposts, GCP Anthos, and Azure Private Stack. Additionally, VMware has established a hybrid cloud offering with each of the Big 3 CSP vendors to further integrate legacy infrastructure with public cloud.
Multi-cloud includes on-premises infrastructure as well as two or more public cloud environments, and generally with connectivity and integration back to the data center (and often CSP-to-CSP as well), as illustrated in the graphic below:
Figure 2: Hybrid and Multi-cloud defined
2.2 Multi-cloud by the numbers
Most enterprises are already operating multi-cloud environments with more than one public Cloud Service Provider (CSP). While many see it as a core component of their forward-looking technology strategy, some have also enabled “multi-cloud by accident” and are revisiting their portfolio and vendor ecosystem to inform the clear path forward (2).
Figure 3: Hybrid and multi-cloud by the numbers
2.3 Why is this becoming an increasing trend?
A number of trends are presenting themselves in the marketplace that are pushing organizations to multi-cloud deployment models, namely with respect to edge computing and a strong desire to maximize the capabilities of the cloud while minimizing vendor lock in. Some highlights include the following:
- Shifting of workloads toward the edge: there is a new demand and need to get closer to the customer, endpoint, or wherever the data is collected for any given use case (3). CSP’s have noticed this trend and are establishing global edge locations, such as Amazon CloudFront for content delivery (4). The demands for fog and edge computing are increasing for a number of use cases that require compute capacity closer to the endpoint, which is increasingly presenting itself with innovation in Internet of Things (IoT) use cases like transportation management and autonomous vehicles
- Increased desire to insource and reduce vendor lock in: IT organizations are seeking to regain control of their products and engineering while commoditizing and reducing focus on core infrastructure where it makes sense, and take advantage of environment and contractual flexibility to the extent possible with workloads and data that are more portable than they were before in legacy environments
- Increased flexibility for distribution: with a desire to promote consistency, portability, reusability of infrastructure, platform and application services
- A rise in containerization: container-based computing, e.g. Kubernetes, has greatly reduced the friction to adopt hybrid and multi-cloud by encapsulating software code and dependencies, enabling software to run uniformly and consistently on any cloud infrastructure
- A strong desire to maximize operational resiliency: namely with Disaster Recovery and backup in the cloud, utilizing cloud as both a DR option as well as leveraging cloud for its native cross-zone and regional redundancy for many core services
- Access to innovation hubs: provided by cloud-native services, namely with modernized development and test resources and tooling and an immediate feedback loop provided by public cloud data and analytics services
- A certain nimbleness and agility with public cloud offerings: that is typically hard to replicate on-premise in the data center at similar speed and scale (e.g. the five classic tenants of cloud computing to include elasticity and scalability, on demand / automated provisioning and self-service, multi-tenancy and resource pooling, broad network access, and measured service consumption) (5)
- The inevitability of cloud, and therefore multi-cloud. Cloud computing has passed the inflection point of being a technical certainty. With such rapid adoption of cloud, more and more organizations are now operating in more than one cloud platform for a variety of reasons — to meet regulatory requirements, diversify risk, avoid vendor lock-in, etc. These organizations want solutions that provide a consistent experience across cloud platforms, giving the edge to startups against native offerings from the major cloud providers. — Ali Ghodsi, CEO, Databricks (Source)
2.4 Why is it hard to be successful with hybrid and multi-cloud?
The trend data above shows that the majority of large institutions have adopted cloud.
- Clients have enabled “multi-cloud by accident”: many have “adopted multi-cloud by accident” with a proliferation of services, as opposed to a deliberate strategy that is increasingly becoming a trend (2) (6).
- Many clients have varying degrees of maturity: across all required components for the end-to-end multi-cloud. This has resulted in a proliferation of services that have varying degrees of maturity, integration with the total enterprise technology ecosystem, and transparency with monitoring and control functions that need to maintain visibility of all resources.
- Orchestration is required across every infrastructure layer: for every cloud environment and layer of the technology stack; there is no “silver bullet” tool or platform that enables the Data and Service fabric required for multi-cloud (see following sections for deep dive)
- Most commodity (e.g., infrastructure) and differentiated services (e.g., serverless) are unique cross-CSP vendor: Many of the core cloud infrastructure services that are fully mature in this stage of cloud and are viewed as commodity still differ across CSPs in nature of pricing, consumption, and product and service model, typically requiring an overlay for multi-cloud abstraction and management (e.g. Cloud Management Platform, multi-cloud provisioning tools such as Terraform, Ansible, Puppet / Chef). Additionally, the more abstracted and newer services such as serverless are even more unique, embedded with the CSP environment, and not easily migrated cross-cloud (e.g. Serverless spanning GCP Cloud Functions, Azure Functions, and AWS Lambda)
- Best of breed capabilities are disparate across CSP vendor environments: according to their respective strength areas, which also frequently change with CSP new service launches and enhancements
These challenges are complex and many are newly introduced obstacles that were not originally experienced in the data center, or even in single or hybrid cloud deployments. However, there is a path to multi-cloud enablement to take advantages of the many benefits that far outweigh the costs.
3 Guiding principles for enabling hybrid and multi-cloud
Figure 4: Guiding principles for enabling hybrid and multi-cloud
Evaluate your current cloud journey to map cost, complexity, and maturity gaps.
Harmonizing on-premise and multi-cloud environment costs, planning and chargeback is difficult, with no “silver bullet” solution available. A mixture of cloud-native financial tools (e.g. AWS Cost Management), vendor tooling and platforms (e.g. Apptio, Cloudability) and custom development for aggregation and establishing a least common denominator and financial transparency can be leveraged. (7) (8)
Mapping the complexity of legacy multi-cloud environments requires next-gen service management tools and processes for end to end orchestration, such as ServiceNow, Flexera, and cloud-native discovery and reporting tools (e.g. AWS Application Discovery Service).
In evaluating current state maturity, many organizations have the foundations (e.g., CSP environment, WAN) but not the end-to-end data and service fabric across all components. Per the following section, evaluate where your environment stands with respect to the standard for Cloud Data and Service Fabric and identify gaps for closure.
Figure 5: Cloud cost, complexity and maturity
Identify what is a priority regardless of platform choice versus what is differentiated.
Typically, as you move “further down the stack” (e.g. core infrastructure such as storage / network / compute) the product specifications, performance, pricing, etc. are generally more commodity and relatively less differentiated (albeit not easily orchestrated per previous section discussion). The further you move “up the stack” with services of increased abstraction and specialization, you generally will find more differentiation and embedding within the CSP for your application, with the positive tradeoff of gaining increased benefits from the cloud. Evaluate each carefully to make the appropriate choices on what is priority for your deployment.
Figure 6: Commodity vs. differentiated cloud solutions
Make balanced cloud architecture choices that maximize control while leveraging the benefits of cloud.
In the same vein as the previous point, there is a continuum of adoption that spans everything from an extreme desire for CSP agnosticism, control, and workload portability, to a full embrace of one or more CSP vendors and everything their environments have to offer. The recommendation is to carefully assess the posture of each adoption model against the broader technology strategy and make balanced cloud architecture and CSP service choices the underpin the vision for your environment and solution strategy.
Figure 7: Multi-cloud adoption continuum
Establish the data and service fabric for cloud integration and orchestration of the total ecosystem.
In order to enable cloud native services in a multi-cloud environment, there is a Data and Service Fabric required to enable all of the best practices discussed previously, to include cross-CSP orchestration and consistency, workload and data portability, and data encapsulation via APIs and microservices. Each layer and component of the fabric collectively acts as the glue and environment abstraction required to “register a service” on the multi-cloud, be it an API, an application service, or any other technology product or service that is exposed to your customers and business partners.
Figure 8: Cloud Data and Service Fabric
Enable an ecosystem promoting innovation and operational resiliency with the right architectural choices.
There is a much broader and separate conversation to be had around the principles of resiliency within the context of cloud and IT infrastructure, including Site Reliability Engineering (SRE), business continuity, and disaster recovery planning.
Specific to multi-cloud, many organizations are leveraging public cloud as a vehicle for innovation to later repatriate to the data center. Take Dropbox as a publicly available example of those who have taken the approach of “getting it right in the cloud”, then later shifting applications, workload and data back on-premise where their new, standardized products can be run on-premise, and in some use cases this can be do so at a reduced run rate. (9)
Figure 9: Cloud repatriation to the Data Center
In conclusion, multi-cloud is clearly the 2020 trend that has either presented itself through organic environment evolution, “on accident”, or most recently by way of deliberate and thoughtful strategic adoption to leverage the advantages of this dynamic delivery and hosting model. Please use these thoughts shared as a snapshot of constructive suggestions and food for thought for best adopting and maximizing the benefits of the ever changing CSP landscape. If anyone is seeing anything differently in the marketplace or pivoting trends toward new and improved models, please do not hesitate to reach out to discuss the rapid evolution in this space that will help everyone keep a collective edge in getting ahead of the benefits of cloud.
Matthew Leybold is an Associate Director in BCG Platinion out of New York City, and is a leader in the BCG Platinion Public Sector and Financial Institutions verticals for Core Technology and Digital. You may contact him by email at firstname.lastname@example.org.
Figure 10: Cloud-native reference architecture (10)
Figure 11: Public cloud service providers are delivering both cloud-based and on-premise offerings that are environment consistent to promote hybrid cloud
Figure 12: US Public Sector considerations, e.g. AWS GovCloud is designed to address specific regulatory & compliance of US government agencies that run sensitive workloads in the cloud
1. BCG: Enabling Digital Transformation with BCG’s Data & Digital Platform. https://www.bcg.com/en-us/capabilities/technology-digital/data-digital-platform.aspx
2. A Clear Multicloud Strategy Delivers Business Value: Get Comfortable With The Reality of Multiple Cloud Partners. Paul Miller. April 18, 2018. Forrester
3. ZDNet. What is edge computing? Here’s why the edge matters and where it’s headed. Scott Fulton III. August 9, 2019. https://www.zdnet.com/article/where-the-edge-is-in-edge-computing-why-it-matters-and-how-we-use-it/
4. Amazon CloudFront. https://aws.amazon.com/cloudfront/
5. SP 800–145, The NIST Definition of Cloud Computing. https://csrc.nist.gov/publications/detail/sp/800-145/final
6. The Enterprisers Project: Multi-cloud: 5 important trends to watch. https://enterprisersproject.com/article/2019/7/multi-cloud-5-key-trends
7. Cloud costs are spiraling — here are 5 steps to get them back under control. BCG Platinion. Oded Kaplan, Rishi Mallesh, Andreas Rindler. https://medium.com/@simphall/cloud-costs-are-spiraling-here-are-5-steps-to-get-them-back-under-control-9df035344e8f
8. Don’t let COVID bust your Cloud budget. Simon Hall, BCG Platinion. https://medium.com/@simphall/dont-let-covid-bust-your-cloud-budget-b94216662278
9. Three years after moving off AWS, Dropbox infrastructure continues to evolve. https://techcrunch.com/2019/06/21/three-years-after-moving-off-aws-dropbox-infrastructure-continues-to-evolve/
10. Cloud-Native Computing Foundation; Cloud Native and Container Technology Landscape: https://www.cncf.io/blog/2017/05/15/developing-cloud-native-applications/
Other references (non-cited or image embedded, please check these out if you want to expand your knowledge further on these relevant topics for cloud and multi-cloud):
1. The Next Platform: The Birth of the Distributed Cloud. https://www.nextplatform.com/2020/02/25/the-birth-of-the-distributed-cloud/
2. 4 Trends Impacting Cloud Adoption in 2020. https://www.gartner.com/smarterwithgartner/4-trends-impacting-cloud-adoption-in-2020/
3. Softchoice. Not Just a Cloud: Perspectives on a Multicloud Approach. https://www.softchoice.com/blogs/advisor/cloud/not-just-a-cloud-perspectives-on-a-multicloud-approach
4. HashiCorp: Unlocking the Cloud Operating Model. https://www.datocms-assets.com/2885/1581718017-com-white-paper-v6.pdf
5. Inside Big Data: Benefits to Consider from a Multi-Cloud Approach. https://insidebigdata.com/2019/07/31/benefits-to-consider-from-a-multi-cloud-approach/
6. Express Computer: Top 7 emerging hybrid cloud computing trends to watch in 2020. https://www.expresscomputer.in/cloud/top-7-emerging-hybrid-cloud-computing-trends-to-watch-in-2020/45504/
7. Data Center Knowledge: The Economic Advantages of Hybrid Cloud. https://www.datacenterknowledge.com/industry-perspectives/economic-advantages-hybrid-cloud
8. Forrester. Assess The Pain-Gain Tradeoff Of Multicloud Strategies. Learn The Difference Between Strategic Multicloud And Ineffective Complexity. by Lauren E. Nelson, Andre Kindness, and Naveen Chhabra. March 19, 2019
9. Forrester. Hybrid Cloud Security Best Practices: Focus on the Five C’s: Console, Configuration, Connectivity, Cloud Data, and Containers. Andreas Cser. June 20, 2019.
10. Cloud Native Architectures: Design high-availability and cost-effective applications for the cloud. 2019, Amazon.com Services LLC. Written by Tom Laszewski, Kamal Arora, Erik Farr, Piyum Zonooz. https://www.amazon.com/Cloud-Native-Architectures-high-availability-cost-effective-ebook/dp/B0788SDV7W
11. Altocumulus: How Corporate IT is entering the multi-cloud — A new formation is rising in the computing skies. The Economist. March 14, 2020. https://www.economist.com/business/2020/03/14/how-corporate-it-is-entering-the-multi-cloud
12. Microsoft’s Hybrid 2.0 strategy: Azure Arc, Azure Stack Hub, Azure Stack Edge explained. ZDNet. November 4, 2019. By Mary Jo Foley. https://www.zdnet.com/google-amp/article/microsofts-hybrid-2-0-strategy-azure-arc-azure-stack-hub-azure-stack-edge-explained/
13. 10 Emerging Cloud Computing Trends To Watch in 2020. CRN. Joseph Tsidulko. November 14, 2019. https://www.crn.com/news/cloud/10-emerging-cloud-computing-trends-to-watch-in-2020?itc=refresh
14. IBM: A blueprint for data in a multicloud world. https://www.ibm.com/thought-leadership/institute-business-value/report/multicloud-data-strategy#
15. Amazon Web Services. Using AWS Cost Explorer to analyze data transfer costs. Ashish Mehra. June 12, 2019. https://aws.amazon.com/blogs/mt/using-aws-cost-explorer-to-analyze-data-transfer-costs/
16. IBM. Tackling APIs in a multicloud world. September 10, 2018 | Written by: Hasan Jilani, Marketing Manager, API Management and Gateways. https://www.ibm.com/blogs/cloud-computing/2018/09/10/apis-multicloud-world/
17. Six Simple Steps Pave the Way to the Cloud. February 27, 2019. By Suruj Dutta , Gitin Grewal, and Hrishi Hrishikesh. https://www.bcg.com/publications/2019/six-simple-steps-pave-way-to-cloud.aspx