Examine critical components of Cloud computing with Intel® software experts
113 Discussions

Why the Network is Foundational for Any Enterprise Multi-Hybrid Cloud Strategy

0 0 1,401

Over the past several years, most organizations agreed that the pace of disruption would accelerate due to technological advancements. McKinsey & Company stated that “industrial sectors will see more disruption within the next five years than in the past 20 years combined” and that was two years before a global pandemic became a profound disruptor in our world.

Companies moved up timelines for digital transformation investments that are focused on:

  • delivering high-speed, high-quality connectivity, applications and collaboration tools to a remote workforce;

  • using the cloud and virtualization to scale services, bolster business continuity, and respond in an agile manner;

  • reimagining customer experiences and interactions in a socially distanced environment; and

  • securing a growing amount of highly distributed data.

While enterprises are investing in these areas of digital transformation and using multi-hybrid cloud as part of their strategy, challenges persist on ensuring their network infrastructure can handle the mounting requirements that digital transformation requires. Now more than ever, enterprises need to build cloud-ready networks capable of securely connecting business locations, remote employees, customer interactions, siloed business applications and data. To address these challenges, Intel is working closely with the ecosystem to help enterprises get the most from their network infrastructure.

Enterprise IT Challenges

According to IDC’s 2019 Datacenter Operations Survey the top priorities and challenges facing enterprise IT organizations across traditional, private cloud and public cloud datacenters were to:

  • ensure security and compliance (44.5%)

  • improve network performance (43.5%)

  • improve flexibility to move workloads as needed (35.3%)

  • improve ability to plan for capacity needs (32.8%)

  • enable internal IT organizations to operate more like a cloud or service provider (32%)

Multi-hybrid cloud is challenging because of the various infrastructure and operational models that often lack flexibility and are difficult to scale. Not only do data siloes make it difficult for company leaders to gain actionable insights, but multiple infrastructure and operating models also make it more difficult to secure distributed applications and data. Related interoperability problems are mounting for enterprises, the ISV community, and cloud and communication service providers alike. The good news is that the choice of network can help address these challenges.

Why Cloud-Ready Networks are Strategic to Multi-Hybrid Cloud

Network performance was a close second place in the IDC survey of top enterprise IT challenges, but I believe it is the key to unlocking progress with the other priorities. Cloud native networks—those that utilize technologies honed in the cloud—are foundational to this. Consider the intersection with the application, cloud management and infrastructure considerations for multi-hybrid cloud operations.

Applications & Workload: Modern applications are complex in nature because they consist of multiple objects in different systems. Adding to this complexity is the fact that enterprises are deploying more applications outside of traditional on-premises datacenters to reside in colocation facilities and emerging edge environments. This trend demands greater sophistication with regards to application latencies, data locations, usage patterns, and compliance and security requirements. When assessing a portfolio of applications, IT organizations evaluate dependencies, connectivity requirements, and current platform support among many other characteristics. Establishing a consistent architecture makes it both possible and more practical to rearrange workload placement based on cost, complexity, and benefit considerations, such as access and data sharing.

Cloud Management: Security, governance, visibility and automation are all important components of a multi-hybrid cloud management strategy. You can’t secure what you can’t see, so reliable connectivity across any cloud is non-negotiable. Modern, cloud networks are prerequisite for network automation, considered an essential ingredient for operating in a multi-hybrid cloud world. Effective automation with artificial intelligence (AI) and machine learning tools requires visibility across the entire network and through all the layers of the network stack in order to detect and resolve issues before they cause outages. IDC estimates that a single hour of enterprise downtime costs an average of $250,000, while some industries, like financial services, can see penalties that run into the millions of dollars per hour.

A software-defined network provides companies with flexibility in order to mitigate cloud vendor lock-in, ensure workload portability, and improve capacity planning.

Cloud Native Infrastructure: A reliable and software-defined network is the first step towards addressing the top, enterprise IT challenges, including improved connectivity, security and compliance for any workload.

The most common approach to network transformation is to deploy commercial off the shelf servers that support full-stack network virtualization across all infrastructure, virtual machines (VMs) and containers. Enterprises are increasing software-defined WANs deployments to connect branch offices, remote sites and applications in SaaS and IaaS clouds and on-premise datacenters. These cloud native networks enable companies to tailor the compute, storage, networking resources for each workload, which supports better performance and scale and lower total cost of ownership.

By creating an end-to-end data pipeline across private, public and edge/branch clouds, enterprises can tackle these common IT challenges.

How VMware and Intel are Enabling Cloud-Ready Networks

Intel works closely with a proven ecosystem of OEMs, ISVs and OSVs, including VMware, to develop optimized solutions that help put data in the right place for the right outcome and at the right price. In fact, Intel and VMware are working closely to enable a consistent, logical architecture that supports automation, centralized management, and flexibility to move workloads freely while supporting application policies.

Gone are the days when an app used to consist of a few VMs and a database. Consider the VMware Virtual Cloud Network that incorporates VMware NSX-T and off-the-shelf servers with Intel Xeon Scalable processors to connect on-premise resources with multiple clouds and facilitate the portability of applications and data with the highest levels of security. For the edge, the cloud-native VMware SASE architecture combines VMware SD-WAN Gateways, VMware Secure Access, the zero trust network access (ZTNA) solution, secure web gateway (SWG) and cloud security access broker (CASB) solutions, as well as VMware NSX Firewall which is VMware’s next-gen firewall functionality. This is all delivered through VMware SASE points of presence (PoPs). These networking and security services can be delivered in an intrinsic or sequenced manner to branch edges, mobile users, campuses, and IoT devices. Intel processors provide a consistent and scalable infrastructure for VMware SD-WAN at the edge and VMware SASE PoPs.

In addition to core network virtualization and connectivity services, such as routing, switching, firewalling and load balancing, the NSX platform includes a set of integrated tools that enhance management, control and security across a multi-cloud environment. With a consistent logical architecture and centralized management, automation and control in place, enterprises can move workloads freely among these resources. The solution also supports consistent policy application and business continuity.

Intel’s platform technologies for the data center, which serve as the underlying hardware in this solution, incorporate decades of application optimizations for diverse cloud, enterprise, and network environments. Intel technologies complement VMware solutions for application connectivity, processing, security, and load balancing across the cloud, datacenter and edge. This spans processors, accelerators, network interface cards (NICS), Intel QuickAssist Technology (QAT), Intel Hyperscan, Intel Tofino™ and other software optimizations.

Given the scale and scope of the multi-hybrid cloud challenges, Intel has been working with VMware to deliver significant performance that supports reliable scaling and high levels of availability from the edge to the cloud. As an example, we recently achieved 1 million SSL transactions per second with VMware Load Balancer running on 2nd Generation Intel Xeon Scalable processor-based servers, and we expect performance improvements with subsequent generations of the Xeon platform.

In addition to load balancing scale, the joint VMware-Intel solution can scale to thousands of hosts today, with a path to tens of thousands, which alleviates the performance limitation facing deployments. These performance improvements and optimizations are intended to provide uninterrupted availability in a cost-effective manner.

Delivering a Shared Vision for Enterprises

For the last decade, Intel and the communications industry have been on a journey to transform the network with a software-defined, agile and scalable infrastructure. Our breadth of technology and depth of experience with a proven ecosystem of partners means that enterprises can accelerate their most ambitious digital initiatives with confidence.

Download a recent IDC white paper titled “Achieving Network Modernization for the Decade Ahead” for more perspectives on the road ahead for enterprises.

My VMworld 2020 presentation with Pere Monclus, CTO of Network and Security business unit, VMware, provides a glimpse into our future plans to break through technical and operational siloes and enable more fully realized digital transformation.

Learn more about the partnership between Intel and VMware.

Written by Bob Ghaffari, general manager of the Enterprise and Cloud Network Division (ECND) within Intel’s Data / Network Platforms Group.

About the Author
Bob Ghaffari is currently Vice President, Network & Edge Group, General Manager, Enterprise and Cloud Division at Intel Corporation. He is responsible for working with key customers and partners to enable best-in-class solutions for the Edge and Cloud with a focus on network and security solution use cases. Ghaffari’s organization works closely with enterprises, service providers and industry partners to deliver agile, secure, scalable and cost-effective network and edge solutions for multi and hybrid cloud implementations. Prior to this position, Ghaffari has held several positions in product development, business development and marketing strategy. Bob has an EECS degree from UC Berkeley and a MBA from Santa Clara University.