A conversation with Mark Shuttleworth, Member OSM TSC, Founder & CEO Canonical Ltd.

Mark Shuttleworth, Canonical

November 2017

The OSM Marketing Work Group recently caught up with Mark to get his thoughts on the role of Canonical within OSM, the accomplishments of the community so far as well as its goals over the next six months.

1. Why was it important for you to be a part of ETSI Open Source MANO?

Standards are vital in the industry, and open source has also come to represent the primary way in which industries collaborate efficiently in technology. From AI to operating systems, the new normal is open source.

 

OSM represents a very interesting interplay between open standards and open source. I expect this will result in cleaner, more implementable standards which benefit all participants. OSM also represents the intersection of cloud, where Ubuntu and Canonical play a very significant role, and software operations, where Juju offers full application modeling and model-driven operations, which greatly simplify the process of rolling out and operating complex topologies of software across the globe. So it is a pleasure to work with the other participants with the goal of enabling rapid and efficient operations at scale, for telcos, on various kinds of cloud.

2. What is your view of the benefits of OSM for operators and software solutions suppliers interested in getting involved?

As we move to software-define services, the challenge is to operate many pieces of software from many different vendors, in many locations, that are tightly integrated, and which need to change all the time.

The pace of change and the complexity of all that software make it impossible to handle with traditional training and operations skills.

OSM aims to provide telco-centric insights and operations that scale, so companies can efficiently roll out new services at scale. It does that by presenting a service view, from the service orchestrator, which behind the scene is driving an application model in Juju, and consuming resources from the cloud via a resource orchestrator.

3. What role do Juju charms play in OSM and other orchestration environments?

Juju captures all the knowledge for application deployment, integration and operations in a package called a charm. Many companies operating the same application with the same charm are effectively sharing all the knowledge, the best practices, and the skills that are encoded in that charm. And because those charms are open source, those companies can collaborate, which means that the cost of operations for that application decline dramatically. Juju is useful anywhere you have large bodies of complex software that are shared by many institutions. In this case, that's the emerging body of Virtual Network Functions (VNF).

Within OSM, the VNF Configuration and Abstraction Layer (VCA) layer is responsible for enabling configurations, actions and notifications to/from VNFs and/or Element Managers (EM). When backed by Juju, it provides the facility to generate generic or specific indirect VNFMs, via charms that can support the interface the VNF/EM chooses to support.

4. The main theme of Release THREE is "Production Readiness". What are some of the highlights in Canonical's contribution that help accomplish this goal?

In Juju we have been working on scale, high availability and performance. We have also been working on the use of containers to reduce latency for workloads that are trusted to run in containers, which is particularly useful for telco apps.

5. What do you see as the key goals the OSM community should aim for over the next six months?

I think the key goal is to get cloud-native VNFs which deliver the benefits of moving to software-defined infrastructure ready to operate with OSM. The more VNFs, the more interesting OSM becomes to operators, which creates a virtuous circle.