The multi-cloud approach is clearly on top of the list for enterprises and cloud experts alike. However, aggregating multiple non-linear, non-coherent, and non-aggregated cloud services into single console to enable seamless data sharing, without compromising integrity, is a big challenge for system integrators and managed services providers. Let’s deep-dive into the benefits of multi-cloud approach, the enterprise-wide changes it warrants, and the role an effective managed services provider can play in making multi-cloud strategy a success
At a time when simplicity is the name of the game, why are enterprises favoring the multi-cloud approach? The answer lies in multi-cloud’s ability to manage the supply-demand challenges prevalent in today’s business environment with flexibility, agility, and on-demand scalability, while enabling cost and performance advantages. There’s another significant advantage to a multi-cloud approach—it affords enterprises freedom from vendor lock-in—a big concern for most of the enterprises.
However, the question is if multi-cloud is a trend that is here to stay or a fad that only the most cutting-edge will embrace? As per IDC, by 2023, the top-4 clouds (mega-platforms) in India will be the destination of choice for 50 percent of workloads, while lock-in will be avoided through multi-cloud and cloud-native approaches to achieve portability. “According to an IDC report, 55 percent of India’s top 500 organizations will have a multi-cloud management strategy in place by 2023, and will include integrated tools across public and private clouds. Though cloud has become a business imperative, enterprises are trying to make sure they truly have choice and there’s no lock-in risks,” explains Vaibhav Gawde, Head, Solution Consulting, Oracle India.
Every industry has its motivation to embrace multi-cloud strategy however the reasons are subjective. On a generic scale, banks and insurance organizations may adopt multi cloud strategy to spread risks for data security and quality; media and entertainment customers may adopt multi-cloud for seamless customer experience and speed-to-market; retail customers would prefer it for low costs and ease-of-access; and manufacturing customers for accelerated delivery and
connected things, elaborates Duraivel Gopal, EVP & Global Head, Cloud & Infrastructure Services, Larsen & Toubro Infotech.
Moreover, customers always had issues with single vendor lock-in or dependency on cloud providers. “With some cloud vendors going through management changes, consolidation through mergers and acquisitions or simply not able to adapt technology effectively for customers, consistent delivery might be an issue. Hence, customers want to ensure multi-vendor strategies. Customers will also look at coexistence of private as well as public cloud and best fitment of IaaS, PaaS and SaaS as per workload requirements,” highlights Sanjiv Khushu, CEO, Evantage IT Consulting Services.
In fact, multi-cloud is rapidly becoming a new model of IT delivery for Indian enterprises as it helps them meet several business goals including streamlined management and agile application development. Mid-sized organizations as well as Internet, media, and gaming customers are adopting the multi-cloud strategy, says Alex Li, General Manager, Alibaba Cloud South Asia.
The adoption of multi-cloud is growing fast in BFSI, manufacturing, automobile, retail and distribution space including e-commerce. “The key workloads where we have seen the uptake are the ones that help the client go digital and embrace self-service, these workloads are primarily mobility, App modernization, extranet portals, and dealer & supplier facing systems,” shares Prashant Sharma, Country Manager, Hybrid Cloud Integration, IBM India/South Asia.
Multi- or Hybrid Cloud?
Multi-cloud refers to leveraging two or more cloud computing platforms, often to meet specific workload needs but without necessarily connectivity or orchestration between them. “Multi-cloud is an IT environment that leverages multiple cloud services from different vendors. It generally includes various cloud deployment models – SaaS, PaaS, IaaS, private, and virtual private. Hybrid cloud complements multi-cloud with a combination of various IT deployment models – on-premise, hosted private cloud, virtual private cloud, public cloud, hosted servers, and Hyper Converged Infrastructure (HCI),” explains Tushar Parekh, MD, Silicon Netsecure.
In reality, the hybrid IT-construct becomes a part of the multi-cloud environment. As organizations start building multi-cloud strategies, they will need to evaluate multiple on-premise, colocated, hosted and cloud options for their applications, data, networks and servers. Depending on workload characteristics and business needs, they would also need to understand public, private and virtual private cloud services required to optimize their IT environment.
It is clear that multi-cloud and hybrid cloud are intertwined concepts with a common goal – of matching workloads with infrastructure, service types, service providers and deployment models. The benefits associated with both models are largely the same – optimize resources, improve performance, reduce risk, lower costs and build agility into IT environments.
“Hybrid multi-cloud architectures will replace the ‘one-cloud-fits-all’ approach. Businesses are seeking to drive value from the integration of public and private clouds by taking the hybrid multi-cloud approach. The approach uses a combination of on-premise, private cloud, and public cloud architecture, with the best-in-class resources from different cloud vendors,” states Sharma of IBM. He further adds that a hybrid approach enables teams to run applications across private, dedicated and multi-tenant public cloud infrastructures, while a multi-cloud approach embraces multiple vendors to support a breadth of enterprise workloads.
The demand for the multi-cloud strategy is driven by many factors, such as the avoidance of vendor lock-ins, increased agility and automation, and the need for a high-level of governance and policies. With the increase in the adoption rate of cloud computing among enterprises, the multi-cloud management market is expected to gain a major traction.
“Apart from cost, quality is one of the key benefits of adopting a multi-cloud strategy. While some of the cloud vendors may be the most sought-after for media and security, other cloud technologies can be good for machine learning and big data, depending on the use-case. With multi-cloud approach, CXOs have the flexibility of selecting the best cloud service depending on the requirement,” notes Li of Alibaba.
According to market experts, following components are essential to have a cohesive multi-cloud strategy that involves a hybrid IT service deployment model:
Comprehensive Cloud Management Platform (CMP): This allows IT teams to manage the entire multi-cloud / hybrid IT ecosystem from a unified interface. A mature CMP will offer extended features such as service design, provisioning, orchestration, auto-scaling, performance optimization and SLA management.
Well-Defined Integration Model: IT teams need to build sufficient integration and interoperability capabilities across on-premise and cloud applications. This can be done using standard (vendor provided) or custom (business need / user-specific) APIs and available integration tools.
Unified Security and Governance Policies: Typically, IT environments do not have uniform security and data governance policies due to multiple reasons, such as discrete systems, siloed operations and absence of an executive-backed GRC plan. A multi-cloud strategy often exposes enterprise data to external systems, and therefore a unified approach to security and data governance is imperative.
Seamless Access by Business Users: For a multi-cloud strategy to be effective, IT teams need to clearly understand critical use cases and business user expectations to build relevant interfaces and provide secure access.
The IT environment for organizations in India is becoming more complex with multiple public clouds, private clouds, and traditional systems needing to be interconnected, integrated, and collectively managed. Most of Indian organizations are clustered around early stages of cloud maturity and find it challenging to move ahead in the adoption curve, discloses Rishu Sharma, Associate Research Manager, Cloud and AI, IDC India.
While the benefits of multi-cloud are well-established, the big question is: how to tie the different environments together to ensure seamless integration across an organization? “Multi-cloud is becoming the new normal, where multiple services from multiple clouds play key roles in customer’s digital transformation. But that also means growing complexity and organizations are challenged to deal with its fragmented nature, increasing complexity, and lack of control when it comes to data, policy, and security,” highlights Trideeb Roy, Head, Data Center Cloud Virtualization, Cisco India and SAARC.
Besides integration, the SLA challenge is also a big hurdle. Maintaining application SLAs become an issue as IT does not have the same control over resources in public cloud architecture as compared to private cloud. Further, multi-cloud SLAs are not standardized as each provider selects their own metrics, restrictions, and exceptions with regards to infrastructure availability. Lack of standards and ambiguity over ownership of responsibilities is a potent mix for disaster—as enterprises cannot easily pin-point which service failed, why, and who is responsible for compensation.
Other challenges include app sprawl, managing multiple compliance requirements and unique vendor portals, ensuring seamless workflow migration across clouds, and maintaining tight security.
Enterprises that are early adopters need to carefully tread their strategy, taking into account the complexity, costs and pitfalls that these challenges might ultimately entail. Costs can multiply in a multi-cloud environment. While at the outset realizing cost efficiencies by leveraging more than one cloud provider might seem plausible, the lack of tools and standards and other inefficiencies to train and track usage and costs across cloud platforms might prove costly in the long run.
The lack of interface standards and portability between different cloud providers creates silos and makes for a challenging and inconsistent proposal when it comes to infrastructure management and developer experience.
Get the Right Tools
Currently, even though most public cloud vendors offer almost similar functionalities, once you use a specific API or service, then your code contains that particular vendor’s code embedded and it is a form of vendor lock-in. You cannot easily move to other public cloud without changing the code. Technologies like containerization, Common Programming interfaces will make it easier to have multi-cloud strategies, says Shailesh Kumar Dave, Vice President, ManageEngine, a division of Zoho Corp.
The good news is that a lot of startups and larger vendors alike have made great progress removing complexity in multi-cloud, and the larger cloud providers have started to be open to interface standardization considerations.
Many of the cloud vendors and MSPs offer CMPs and proprietary tools to help enterprises effectively manage a multi-cloud environment. They consolidate non-linear, non-coherent, and non-aggregated cloud services into single console and enable data sharing across multiple cloud platforms, without compromising on integrity.
The right cloud MSP brings the expertise to: assess workloads, match workloads to cloud types, ensure application readiness and glitch-free migration, and finally supervise execution through managed services delivery. The goal is to enable IT teams to operate seamlessly across cloud platforms, easily managing, securing, governing, and self-provisioning workloads.
Ideally, a CMP must have the following traits:
- A CMP enables organizations to manage their entire cloud estate from a consolidated viewpoint. Integrating with technology from the leading cloud providers, it provides total visibility into all cloud resources to enable companies to maintain security, control costs, and ensure governance across multiple accounts and providers.
- One of the most desired features of a CMP is cross-cloud tagging.Tagging capabilities enable a company to tag assets throughout its entire environment, regardless of which or how many cloud providers it uses. Tagging, when paired with analytics, gives a company visibility into who’s using which cloud assets and shows how those assets are being used and why. Such visibility is critical for effectively managing total cloud spend, as well as deploying standard policies and controls across the assets.
- Another CMP capability that’s especially important in a multi-cloud environment is its ability to discover resources.Cloud makes it easy for essentially anyone with a credit card to go online and spin up a new resource. That’s a blessing and a potential curse. Without being able to discover resources, it’s easy for the cloud equivalent of “shadow IT” to proliferate. An ideal CMP mitigates the proliferation of shadow IT and is able to understand assets from a scale perspective at all times. A CMP enables a company to scan its entire cloud estate to find out what’s being added and assume control over it—which helps control costs, as well as ensure compliance and security.
A report by IBM highlights that by 2021, 98 percent of organizations plan to adopt multi-cloud architectures, but only 41 percent have a multi-cloud management strategy and just 38 percent have procedures and tools to operate a multi-cloud environment. “So we will see IT organizations undergo a cultural shift in the way they work and require teams equipped with new skill-sets. The approach will create job titles such as cloud service broker, cloud architect and cloud automation engineer, and it will be increasingly commonplace to manage the changing environments,” foresees Sharma of IBM.
AI defines the future
Hybrid-cloud and multi-cloud are becoming the new norm. There isn’t a “point solution” for these, and there might never be a single all-encompassing solution. Rather, it is a holistic and integrated approach in an enterprise’s overall cloud strategy. As the support and standards in this space evolve, the current challenges should wane rapidly, leading to progressively quicker adoption.
AI holds tremendous potential for powering the next wave of innovation in a multi-cloud world. For instance, a next-generation multi-cloud AI platform will be intelligent enough for workloads to self-orchestrate themselves and deploy programs developed using Machine Learning (ML) algorithms across multiple nodes, provisioned on-demand.
With tech behemoths such as Google, Amazon, and Microsoft leading the way, many Platform-as-a-Service (PaaS) solutions have started incorporating AI capabilities. Cloud AI services and platforms enable enterprises to leverage AI or cognitive computing capabilities via simple API calls, eliminating the need to invest in sophisticated AI infrastructure. The new era of the AI-first cloud is upon us and forward-looking organizations are prepping for new computing infrastructure to support next-gen cloud programming paradigms and frameworks.
Multi-Cloud: Key Advantages
In this increasingly complicated environment, enterprises are realizing that one cloud is not sufficient to meet all their myriad needs. Different workloads achieve optimal performance and cost utilization on different cloud platforms and a multi-cloud strategy enables enterprises to create the optimal solution from various best-in-class technologies and services. Many of the enterprises favor AWS for features, scale, and ecosystem, while those that are Microsoft-centric opt for Azure. Google Cloud Platform is the go-to choice for enterprises when better storage, technology and network are their key requirements.
“The multi-cloud approach makes it easier to switch between cloud service providers to leverage on better pricing or technical capabilities offered in a cut-throat competition. The idea is to establish a system to run workloads within the most appropriate environments – whether the goal is either capitalize on performance optimization or take commercial savings into account,” says Nippun Gupta, CEO, The Fourth Dimension Consulting Group.
Here are the three key advantages of using a multi-cloud strategy:
High availability, low latency: Multi-cloud works perfectly with edge computing. Modern businesses and consumers expect real-time availability, and the latency that is inherent in cloud services delivered from remote servers is simply unacceptable. With a multi-cloud infrastructure, the datacenter at the edge, i.e., the one closest to end-users can serve the requested data with minimal server backflows, enabling a seamless and unified end-user experience.
Superior capacity planning with on-demand flexibility and agility: To address frequent spikes in business and end user traffic, organizations need a load-balanced environment that can support automated scaling while enabling on-demand flexibility to access either a public or private cloud infrastructure, without any IT intervention. A multi-cloud approach offers scalability and throughput that far exceed traditional options.
Superior cloud native development: Cloud native development and operations patterns are enabling an organization to adopt a multicloud strategy. Cloud native development enables organizations to build a portable software stack that is DevOps driven, free from vendor lock-in and capable of delivering a superior set of capabilities than can be gained from a single cloud.
Reduced risk of service disruption: As IT becomes the center of competitiveness for nearly every enterprise today, companies have to think about being able to survive a widespread outage or security vulnerability. Multi-cloud operations spread the risk of service disruption across a larger surface area. While public cloud providers such as Amazon, Google and Microsoft offer very robust and reliable services, there are times when having multiple clouds and providers are essential. It’s not often that these cloud providers have outages, but when they do, it makes national news. Enterprises feel bereft without access to applications, servers and most importantly, their data.
Moreover, many of the enterprise IT teams and executives may overlook a clause that enterprises are responsible for their own data protection. Public cloud providers are not responsible for your data. The service level agreements (SLAs) may be 99.99 percent, but that is only for network availability and the durability of the infrastructure, not customer data and data availability. Data falls under what is often termed the shared responsibility model.