Into the Cloud – Part I
Part I: The heart of Cloud network architecture
The need to boost the Cloud Continuum experience is pressing and immediate. Read on to know how:
Tech consulting giant Accenture recently wrote: “Not long ago, the definition of a cutting-edge IT network was a fleet of static desktop PCs with a robust corporate data center as its beating heart.
Today, that seems … almost quaint.”
An entire continuum of cloud solutions is taking over, as countless laptops and mobile devices continuously connect to services like Microsoft’s Office 365 and Salesforce.com or applications on AWS, Azure, and Google Cloud Platform. A growing array of sensors from connected machinery and edge devices is churning out data while a new army of remote workers needs flexible “anywhere, everywhere” access.
The Heart of Modern Cloud Network Architecture
This need for “anywhere everywhere” access is what birthed the cloud in the first place – forcing today’s networks to meet a wide and dynamic set of needs for seamless data, application, and platform connectivity. These aren’t things that can be dealt with a quick fix – they require a complete reconfiguration of the network architecture. This entails:
- Building a secure network foundation for cloud services:
The first step is constructing a solid, safe foundation, i.e. rigorous security that can cover a wide and complex web of traffic. Data centers that once housed network servers are now to be moved to the cloud for greater integration. The wireless nature of 5G technologies is set to make device mobility easier. A configurable architecture, at such a point, is now prepared and dynamic enough to add new cloud-native projects.
- Preparing for a multi-cloud/hybrid environment:
Not all organizational workloads are set to be hosted on the same cloud, and a multi-cloud/hybrid architecture is inevitable. Therefore, hosting a multiple/hybrid cloud system needs to be supported by a dynamic network.
Accenture writes: “Rather than treating networking as a problem to be solved (and re-solved) in each new cloud project, the goal is to provide flexibility so organizations can choose options that meet their unique needs—whether that’s machine learning services on the Google Cloud Platform, a data lake in Azure or workloads in AWS. To simplify this process, consider using a software layer product to manage the WAN and 5G connectivity across multiple clouds.”
- Design resilience is key:
“The best networks are those built with the idea that they’re going to fail.”
Redundancy must be built into a network from the start, to create resilience across the whole system. No single outage should, at any point, impact users’ abilities to access services.
Netflix’s ‘chaos engineering’ model, for example, uses a tool called ‘chaos monkey’ that causes random system outages in order to test resilience. Enterprises can then address weaknesses and strengthen overall systems and confidence in delivering round-the-clock availability.
- Automation to be central:
As opposed to the static environments of networks of the past, cloud-based services spin up, at given time periods, thousands of workloads that cannot be handled by manual receivers. Instead, a highly automated programming network layer needs to be integrated with these cloud workloads and enterprise tools via a varied set of APIs. This helps in unifying network monitoring as well as analysis, thereby accelerating a network’s ability to flex business needs.
“At a large oil and gas company, Accenture took over operations of the client’s global network and deployed more than 125 pre-built use cases for different automation scenarios to quickly reduce operations team activities. Using a variety of tools, Accenture customized that automation into the production environment. As a result, the company saw an 80% reduction in incidents.”
- Building new ways of working:
Organizational goals should be to allow cloud service access from ‘anywhere and everywhere with the right security, bandwidth, and performance, and at the right costs. The Secure Access Service Edge (SASE), for example, integrating security capabilities and SD-WAN is an increasingly lucrative option.
(To be continued)
Know more about the syllabus and placement record of our Top Ranked Data Science Course in Kolkata, Data Science course in Bangalore, Data Science course in Hyderabad, and Data Science course in Chennai.
