Day One: NorthStar Controller Up and Running

Day One: NorthStar Controller Up and Running is intended for all networking professionals working on WAN and LAN environments that make use of IP/MPLS services. It should also be of interest to high-level technical professionals looking to understand the fundamentals of Juniper’s approach to Software Defined Networks (SDN) in MPLS transport networks.

The book introduces readers to the Juniper NorthStar Controller by focusing on the discovery and visualization of IP/MPLS networks including the ability to visualize the paths different LSPs take on the network, monitoring the status and utilization of the network in real-time, and, modeling the impact of network changes, among other use cases.

“Modern networks are increasingly in need of more sophisticated traffic-engineering services while at the same time the management of even the most basic traffic engineering requires simplification. The Juniper NorthStar Controller provides the necessary tools to deliver just that! With it’s programmable and open framework, it offers Service Providers, Content Providers, and Enterprises a simple, yet sophisticated toolset to take control of their network by optimizing their existing infrastructure while driving new services.”

- Colby Barth, Distinguished Engineer, Juniper Networks, Inc.

Sample Pages

Download Book

Day One books are a free download for our J-Net members*. If you’re not a J-Net member, create a user account now. It’s fast and there’s no commitment or spam. Once you’re a member you can come back and download any of the Day One books.

Download: J-Net Member
Download: Not a J-Net Member

* If you have an existing Juniper user account, you can use it to login to J-Net

About the Author(s)

Patricio Giecco is a Product Architect who has worked in a number of positions in the Networking industry, including Solutions Engineering, Technical Marketing, and Product Management. In his tenure at Juniper, Patricio has written many application notes, co-authored Junos Security (O’Reilly Media, 2010), and was a recipient of a Juniper Star award in 2011 and 2014.

Author Q & A

What got you started on this book?

I started playing around with the NorthStar controller in its early days, as I was working on a project dealing with service provisioning. The idea of a centralized controller isn’t new, and the industry has gone though a few iterations of this idea using different underlying technologies. The difference is that, as I see it, this time we managed to strike a balance between what must be centralized and what can be distributed. By leveraging years of experience in IP/MPLS networks, we allowed for centralized control of aspects of the network that can benefit from it (such as the placement of LSPs), while maintaining distributed capabilities when central control is not required. Maintaining distributed control in the network is also crucial as backup alternative when problems in the centralized controllers occur.

Software defined networks (SDN) allow for a great degree of flexibility, but one should not ignore all the mechanisms developed over years of experience in large scale networks (such as LSP protection mechanisms, auxiliary protocols used to detect faults, troubleshooting mechanisms, support of in-service software upgrades with minimal disruption to network traffic, etc.). Leveraging the built-in capabilities of IP/MPLS networks, while allowing the delegation of some network control to centralized controllers seemed to open new possibilities, some of which lay dormant for years as we developed fully distributed solutions.

After this (perhaps obvious) realization, its was only natural for me to get involved in the project. I started working on a short application note detailing some of the options of how to connect a NorthStar controller to a network. As most engineers, I always try to understand the underlying mechanisms and motivations of any technology; given the flexibility that NorthStar affords, the application note kept growing. After a short conversation with the editor, we decided it would be good idea to turn it into a Day One Book.

Who is this book for?

This book is for network engineers interested in software defined networks (SDN) and how some of the ideas behind SDN are applied in wide area networks. The book can also serve as a quick reference to anyone looking to understand some of the fundamental concept behind SDN in the WAN, so readers that aren’t interested in the low-level details of the Juniper WAN SDN solution can still get an overall picture of its applicability and main characteristics.

After reading this book, what’s the take away?

There are probably a few and it depends on what each reader is looking for. For me (as I was writing the booklet) was to think on how some of the SDN ideas were translated into existing Wide Are Networks, without throwing away what made MPLS so powerful. This is, perhaps, the biggest difference between SDN in the WAN and in Datacenter environments. The solutions developed in datacenter environments mostly focused on providing basic connectivity. It wasn’t until recently that other use-cases such as the connection of virtualized network functions or, more generally, the interconnection of a large number of virtual machines highlighted the need to orchestrate both the network and other parts of the datacenter infrastructure. With it, the need to provide end-to-end traffic control, orchestration of VMs and the network and, in general, the need to manage the network through a set of programatic interfaces gave birth to SDN and the set of underlying protocols used to achieve this goal.

Because some of these requirements were already solved in wide area networks, the ideas and concepts brought forward by SDN were not applied in the same way. I hope readers will start thinking about how some of the SDN concepts being developed for datacenter networks are translated into the WAN (and vice versa).

What are you hoping that people will learn from this book?

The book focuses on the details of how to connect and configure a NorthStar controller to manage an IP/MPLS network. Technically inclined readers will learn about the different protocols and mechanisms used to control wide area networks, including:

  • The Path Computation Element Communication Protocol (PCEP)
  • The use of BGP-LS to convey traffic engineering information
  • How IGP protocols can be used to discover topology and traffic engineering information

Finally, although the book only provides an introduction, some of the use-cases that NorthStar solves are also shown. Curious readers will be able to discover some of the more advanced use-cases on their own, after the basic scenarios are presented, including:

  • Network topology discovery
  • LSP discovery and visualization
  • Provisioning of LSPs through the controller
  • Optimization of LSPs

What do you recommend as the next item to read after this book?

That’s hard to say, as I normally end up lost in a web of documents, papers and standards every time I start looking into a new topic. Depending on how deep one wants to go, some of the interesting reads are:

  • RFC5440, the PCEP protocol specification
  • RFC4655, a Path Computation Element (PCE) -Based Architecture
  • Day One: Understanding OpenContrail Architecture - this Day One Book will be interesting for readers that want to understand how SDN is implemented in datacenter environments, and the use-cases it addresses

What’s your inspiration?

Other than a borderline obsessive interest in networking in general, as most engineers, I love to understand how things work. From fundamental physics (which probably would be my next career of choice, if I were to go back to school) to applied technologies in out day-to-day lives. There are few things as gratifying as discovering new things, learning new ideas or solving problems (even simple ones) when applying new and long learned concepts. Perhaps more importantly, my wife is also a constant source of motivation; I can always trust her to point out my failures in understanding, or put difficult concepts in simple terms. She manages to work long hours as a Professor of Engineering, raise a child, and deal with a stubborn husband providing a display of grit rarely seen.

What’s your favorite bit/part in the book?

I think chapter 3. Because after all the work required to set things up, readers get to play with the controller and get a glance at what it can do. Playing around, testing new scenarios or exploring the device and system logs is, at least for me, the best way to assimilate some of the concepts.