Sorry, you need to enable JavaScript to visit this website.

Hybrid Multicloud as a Future

Hybrid Multicloud as a Future
July 16, 2020

Co-authored by: Abhirup Chakrabarti

Cloud, an almost ubiquitous storage medium today is lauded for its agility and flexibility. Technically, Clouds are a combination of multiple data centers comprising network connected compute and storage resources. These resources are shared and virtualized to make them work for everyone and can be dynamically scaled on demand. Consider it as a smart grid, where irrespective of the input resource, whoever is connected to the grid gets electricity. Consumers only pay for the services they use. Every business is directly or indirectly impacted by the adoption of cloud to create applications and deliver services. Cloud computing has created new opportunities for both businesses and consumers. According to IDC, the total volume of global data will reach 175 zettabytes by 2025. Companies need to bank on that data to stay competitive in the market.

The Emergence of Hybrid Clouds

With a myriad of cloud service offerings from public providers like AWS, Microsoft Azure, Google Cloud Platform, Alibaba Cloud and IBM Cloud (IBM recently acquired Red Hat to disrupt cloud market), companies are well-positioned to embrace the adoption of cloud in the foreseeable future, and realize the multifarious possibilities of cloud computing. Public cloud services is not the entire picture though – companies adopt private cloud, which constitutes hosted cloud services (involves third party vendors) and on-premise services based on the business priorities. Public cloud gives economies of scale the benefits of scalability and computational capacity whereas private cloud gives flexibility, improved resource utilization (virtualization), IT security benefits, and on-premises provide ownership along with full data control. Some of the use cases of public cloud services are customer-facing applications like email, CRM, HR management, web applications, and so on. Private cloud finds its place in R&D, manufacturing and supply chain, compliance data, ERP, and some unique cases of on-premise can be factory floor applications (low latency requirement), legacy applications, and workloads involving critical and strategic data.

Organizations benefit from the above configurations on different instances ending up using the combination of the same, enter the hybrid cloud. Hybrid cloud is an enterprise-level IT architecture that uses different infrastructure environments like public, private, and on-premises. It offers the best of both worlds from public and private setups giving both scalability and control. The key area of focus here is interconnectivity and interoperability, how data is seamlessly transferred between separate clouds. All the individual pieces from different entities are orchestrated to result into a unified business outcome. The development of an application can start in a private cloud to have more control, and deployment can happen in a public cloud to scale it for widespread usage. ‘Cloud bursting’ can be realized to effectively transfer data from private to a public cloud when the workload is high and computational power is in demand. The result is the optimization and better utilization of resources at disposal.

Hybrid Multicloud in the Making

With hybrid cloud established, multicloud by contrast is how an enterprise uses multiple cloud providers to cater to different business needs. Hybrid cloud combines benefits of a public and private cloud but the answer to the question as to why an organization needs multicloud strategies, lies with the flexibility and ability to limit interoperability. It gives companies an option to use more than one public provider like AWS or Azure for a unified approach to benefit through sharing, or a siloed approach to protecting data. Companies can avoid vendor lock-in, where they are fully dependent on one vendor and can use other vendors for their customized offerings based on vendor strength areas. Cloud providers are accepting the fact that companies cannot be forced to stay with a single provider. Different business functions in an organization can use different cloud platforms, and cloud providers are catering to that by managing workloads with hosted applications on a different competitor platform.

With the standalone benefits of a multicloud setup, they can be hybridized with other public or private clouds for better visibility and utilization of computing resources giving rise to the reality of hybrid multicloud computing. The flexibility to choose which offering to use (from any public provider) available in the cloud market based on organization needs can amplify the already existing cost efficiencies of the cloud. According to IDC, more than 90% of companies will adopt hybrid and multicloud strategies by 2024.

“Hybrid multicloud solves the underlying business challenge of making the most from cloud investment and will pave way for companies to adapt in the ever-changing digital landscape.”

This tantalizing prospect of hybrid multicloud is luring public providers like Amazon, Microsoft, Google, IBM to bet huge on its custom offerings. The bigger piece of the puzzle has always been managing the on-premise workload of companies. To realize the vision of a truly hybrid cloud, cloud providers are venturing into exactly that by offering a means to run applications on on-premise locations with the help of cloud computing services. Touted as ‘Hybrid cloud as a service’, the early adopter was Microsoft with its Azure Stack followed by other providers like Amazon AWS Outposts, Google Cloud Anthos, and Red Hat OpenShift.

Going Cloud-Native

The larger motive of adopting hybrid multicloud is to enable enterprises to move from on-premise to cloud-native computing. Other than using the public cloud for storage and processing, enterprises can benefit from the same underlying technology that enabled the cloud providers to develop their applications. Cloud-native applications are developed and deployed using services and infrastructure in the cloud itself. This enables enterprises to use the full potential of cloud computing and reimagine the way applications are built. Cloud-native applications use modern technologies like Microservices architecture, Containers (Docker, Kubernetes), Agile, and DevOps. Applications are dependent on the cloud, which makes them portable and scalable across multiple platforms ensuring continuous delivery.

How Google Anthos rises to the occasion

Anthos is Google cloud’s multi-cloud and hybrid application cloud platform designed to meet the paradigm changes brought forth by hybrid and multicloud. In a nutshell, it provides a consistent platform for all application deployments, which can be managed across multiple clouds (public or private), on-premises, or edge resources. It can be envisaged as a modern application platform that enables enterprises to focus on delivering innovative applications using the cloud instead of managing the complex pieces of it.

Cloud-native applications are developed and deployed using services and infrastructure in the cloud itself.

Powering the Anthos platform is the Google Kubernetes Engine (GKE), Google’s container-based service. Anthos provides GKE on GCP cloud and GKE On-Prem for on-premise data centers managed by the enterprises to develop and scale applications as they would do on the GCP platform. API’s and services outside the application are managed by Istio, an open-source service mesh developed for microservices. Anthos Migrate can enable customers to easily shift from virtual machines (VM’s) on other clouds or on-premises to containers in GKE. Cloud Interconnect from Google ensures a reliable connection between cloud networks. GCP marketplace can also be leveraged as part of Anthos. Anthos aims to bring the same GCP experience as in Google Cloud to on-premises. With holistic data sets from multiple clouds, businesses can leverage Google’s AI, ML, IoT, and data analytics capabilities to draw deeper insights. Anthos differentiates itself from competitors like AWS Outposts or Azure Stack by using the existing hardware on-premise and from vendors (VMware, Intel, HPE), extending support for other public clouds (AWS, Azure), embracing the open-source cloud platform (Kubernetes, Istio). It aims to revolutionize the cloud-native application ecosystem to build, deploy, deliver, and manage applications anywhere.