Today, we talk to network architect Andrey Khomyakov, who s helped his organization make the transition from traditional to open networking in production.

What s it like to disaggregate switch hardware from network operating system? Was it hard? Does it really work? What s support like? Was any money saved? Will I get fired for suggesting my company go the open networking route? Andrey will answer these and more questions.

Pete Lumbis, long time friend of the Packet Pushers and contributor to the networking community, also joins with some of his insights learned in his role at Cumulus Networks. This is not a sponsored show. We just want to share the real-world experience of going new-school disaggregation with your networking.

We’ll look at the distinctions between open networking, disaggregation, and whitebox. We’ll also look at day-to-day operations with whitebox, common problems, how to choose a whitebox, and more.

Today on the Priority Queue, we re gonna hide some information. Oh, like route summarization? Sure, like route summarization. That s an example of information hiding. But there s much more to the story than that.

Our guest is Russ White. Russ is a serial networking book author, network architect, RFC writer, patent holder, technical instructor, and much of the motive force behind the early iterations of the CCDE program.

The latest tome to flow from his keyboard (and mine, actually) is Computer Networking Problems and Solutions available on Amazon right now. While I wrote or contributed to several of the chapters in this book, Russ did the lion s share, and we re going to dive into one of his book chapters, devoted to the topic of information hiding.

We discuss the reasons for information hiding in the control plane, including resource conservation and reducing the failure domain; the pros and cons of dividing a network in multiple failure domains with information hiding; and the criticality of convergence.

We also talk about techniques for information hiding, including filtering reachability information and using overlays.

Sponsor: InterOptic

InterOptic offers high-performance, high-quality optics at a fraction of the cost. Find out more Continue reading

You are a problem…maybe the biggest problem of all. No? The crashing router code is the biggest problem? The leaking memory in the switch?

The app needs layer 2 stretched between data centers–what problem could be worse than that?

Today on the show, we re here to argue that, no…it s you. And me. And everyone else you work with.

With us today to defend the idea that technology problems are really people problems is Eyvonne Sharp, network architect and co-founder of The Network Collective.

We talk about how people and processes can contribute more to a problem than a technology. We also talk about three different organizational culture types (Pathological, Bureaucratic, and Generative), how to evaluate your own organization, and Eyvonne recommends a few books on team building and culture development.

Show Links:

Eyvonne Sharp on Twitter

The Network Collective

Using the Westrum typology to measure culture – Andy Kelk

Forget about broad-based pay hikes, executives say – Axios

The Undoing Project – Michael Lewis

The Five Dysfunctions of a Team: A Leadership Fable – Patrick M. Lencioni

Team of Teams: New Rules of Engagement for a Complex World – General Stanley McChrystal

One promise of hyperconverged infrastructure (HCI) is ease of management. Break down the silos, put all the components into a unified whole displayed on a single pane of glass, and voila! Apps are served.

But networking hasn t been integrated as effectively into HCI as the other components of the IT stack. Networking, even in an HCI world, tends be difficult. And with the dynamic needs of HCI, networking just isn’t keeping up.

The days of standing up the network and letting it run are past, because a best effort, rough approximation of how the network should behave isn t something you have to settle for anymore.

Discussing integration of HCI with networking is Big Switch Networks, our sponsor for today s Priority Queue. Prashant Gandhi, Chief Product Officer at Big Switch, is our guest.

We talk about why “best-effort” networking isn’t suited for HCI, and look at HCI-specific operational issues and use cases including container networking and multi-tenancy.

For hands-on experience with Big Cloud Fabric, register for BSN Labs, a demo environment in the cloud that lets you experience the technical differentiation, management CLI, and GUI of Big Cloud Fabric.

Show Links:

Solution Brief: Scale Out Networking Continue reading

The Internet is a network of networks. Where do each of these networks meet to form the global Internet? At Internet Exchange Points or IXPs. In North America, these IXPs are also known as network access points, or NAPs.

Over the years, connecting to a NAP has become increasingly crucial for service providers to get right because of the sheer volume of traffic the Internet carries these days (Hello, Netflix!), the complexity of service provider peering agreements, and endless troubles with security threats.

Joining us today to discuss how to better plan, design, operate, and secure peering is our sponsor Cisco. Our guests from Cisco are Phil Bedard, Service Provider TME; and Bruce McDougall, Consulting Systems Engineer.

We discuss the evolution of Internet traffic flow and interconnection, how peering designs among service providers have changed, the role of telemetry and data, and peering security issues.

Show Links:

BGP Monitoring Protocol (BMP) – IETF

Internet Edge Peering – Current Practice – GitHub

BGP Operations and Security – IETF

Observing BGP activity with BGP Monitoring Protocol – Cisco

Streaming Network Analytics System (SNAS) – snas.io

The Death of Transit And Beyond – Geoff Huston (PDF)

Eyeball network – Wikipedia

Today on the Packet Pushers Weekly show, we investigate how to monitor hybrid and SD-WAN.

If your WAN looks like a mix of legacy MPLS, SD-WAN, and uplinks to cloud, this is your show. Our sponsor today is LiveAction, who is going to shine a light on the hybrid and SD-WAN through monitoring and automation.

Our guest is John Smith, Founder, CTO and EVP of LiveAction.

We talk about LiveAction’s software and how it works, why it’s essential to have visibility into your hybrid WAN and SD-WAN, and how LiveAction can provide highly visual and intuitive insights and actionable intelligence for day-to-day operations, troubleshooting, and long-term planning.

Show Links:

LiveAction’s Packet Pushers Resources – LiveAction

LiveAction on Facebook

LiveAction on Twitter

LiveAction on LinkedIn

LiveAction on YouTube

LiveAction on Google+

Today on the Weekly show, MPLS in the enterprise, especially for the use case of segmentation.

Should you do it? Why would you do it? Considering the equipment you ve got, could you do it even if you wanted to? And even if you should, you would, and you can, what about all those other people you work with?

Discussing this emotionally fraught topic of MPLS in the enterprise is Tom Ammon, a senior network engineer who s supported it in a few different environments. He blogs at blog.tomammon.net.

We start by discussing the general case for MPLS in the enterprise, walk through the hardware and software requirements necessary to support MPLS, and then drill into segmentation more specifically.

We also look at reasons why you’d want to use MPLS for segmentation, and explores issues around operations, skillsets and training, and troubleshooting.