What Makes an Environment Cloud-Native? Virtualization vs. Cloud Computing

Posted by Scott Sumner on Thursday, November 30, 2017 with No comments

The increasing attraction of virtualization and cloud computing technologies has pushed cloud- native environments further into the spotlight. Here, Michael Rezek, VP of Business Development and Strategic Partnerships at Accedian, explores the what, how and why of cloud-native.

What is a cloud-native environment?
MR: It starts with virtualization—the first step to cloud computing, which separates infrastructures through the use of software.

Once infrastructure is virtualized, it provides access to shared pools of configurable resources, such as servers, storage, or applications, which are provisioned with minimal management effort. Only after all component infrastructures in a system are virtualized does the environment truly become “cloud native.”

What characteristics must cloud-native environments possess?
MR: Simply having a virtualized infrastructure or virtualized software application does not equate to being cloud-native. According to the National Institute of Standards and Technology (NSIT), a cloud-native environment should possess all of the following characteristics:
  • On-demand service
  • Broad network access
  • Resource pooling
  • Rapid elasticity
  • Measured service or pay-per-use model

Is ‘cloud-native’ merely a hyped-up concept, or a proven technology?
MR: There is a great deal of hype surrounding cloud computing and virtualization, and coincidentally cloud native environments, with many wondering about their legitimacy as proven technologies.

In some ways, cloud-native is in danger of becoming seemingly unattainable since even the smallest dependence on hardware can disqualify an application from being designated as such. However, it’s important to remember that virtualizing functions is a gradual process and won’t happen in one fell swoop.

Cloud-native infrastructures and cloud-native applications are proven technologies, with the former successfully deployed in data centres globally, while the latter are foundational to software-as-a-service (SaaS) models.

Cloud-native applications—or cloud-native virtual network functions (VNFs)—is where it gets a little complicated. While cloud-native VNFs do exist, their successful transition to a virtual network is still being proven. The challenge lies in the apps (VNFs) not necessarily being centralised in a data centre but instead spread across the network’s distributed “web” with numerous endpoints. This compromises the resource-pooling element of a cloud-native environment due to the sometimes limited pools and clusters of resources at remote points of the network edge, or at aggregation points.

What are some benefits benefits of cloud-native environments?

MR: Three benefits in particular stand out, and they’re all focused on networking and computing happening automatically:

  • Auto-provisioning – in a telecom use case, operators can leverage a cloud environment allowing customers to self-serve their applications without the need for additional resources to deploy new services.
  • Auto-scaling – operators do not have to pre-provision purpose-built networking equipment manually, but can instead use software orchestration to automatically spin up and tear down compute resources, according to customer demand.
  • Auto-redundancy – redundancy can be automated by leveraging pools and clusters of compute resources along with a redundancy policy
What are some challenges with cloud-native environments?

MR: To reap the significant benefits of cloud-native environments, operators must overcome challenges in several key areas:
  • Organizational: Historically, IT designed and managed the compute infrastructure, while network organisations were responsible for designing and operating the network. However, in a cloud-native environment, these two distinct domains must morph into one cohesive unit—leveraging each other’s skills, and cross-training one another to design and operate a holistic virtualized network environment.
  • Financial: Telecom companies have invested billions of dollars in purpose-built networking infrastructure. Migrating to a cloud-based physical infrastructure with software licences will therefore be a costly undertaking. Indeed, the cost may create a barrier for smaller operators wanting to migrate their infrastructures.
  • Network management: successfully managing a software-based network consisting of numerous VNFs is no easy feat. Operators will need to oversee the integration of open APIs to each VNF for management and control. The complexity of working with different standards in an open environment is especially difficult.
How would you sum up the future for cloud-native environments?

MR:
Communication service providers (CSPs) are under mounting pressure to transform their systems and infrastructures to become more agile, able to deliver services at a push of a button. Failure to do this will not only see them struggle to keep up with the competition, but potentially lead to their demise. For today’s telco, the answer lies in the cloud, and specifically in cloud- native environments—which, if implemented correctly, can boost network efficiency, reduce expenditure, and enhance quality of experience (QoE) and quality of service (QoS) for subscribers.


Categories: , ,