Heavy Networking

To run AI workloads, a network needs thousands of GPUs and those GPUs must operate in sync. If there is congestion or dropped frames, very expensive efforts could be delayed or disrupted. While there are advantages to using Ethernet for AI networking (including engineers well-trained in the protocol and a robust ecosystem), it wasn’t designed... Read more »

Where there are containers, there is networking. Today we dig into the networking that underlies Kubernetes, the open source orchestration platform for container-based applications. Our guest Karim El Jamali takes us through the essential concepts: Nodes, pods, clusters, CNIs, virtual ethernet pairs, ingress controller, eBPF, and service meshes. As container-based applications grow in popularity, it’s... Read more »

Fiserv is one of the largest payment processors in the world, In 2023 it handled more than 35 billion transactions worth $2.03 trillion US dollars. Its network is critical to the business. The organization knew it needed network automation, but early attempts got some things wrong. On today’s Heavy Networking we talk about how Fiserv... Read more »

Matt Horn built a data center network through automation, remotely. This is the future of network engineering. Matt shares how his team did it technically: Terraform, a little Ansible, leveraging pipelines, etc. But he also shares the processes and culture that made it happen: Management and peer buy-in, tight enforcement based on user access, and... Read more »

Today we metaphorically pop open the hood of switches and routers, taking a look at the mechanics of how they work. We cover the three states: configuration, operational, and forwarding. We talk RIB and FIB, along with CAM, TCAM, and MPLS. We also cover line rate, port-to-port latency, and buffers. Whether it’s been awhile since... Read more »

Right now, we have the building blocks for network automation, but we don’t have end-to-end designs or complete systems. It’s like having a bunch of Legos but no instructions for how to build your spaceship. Ryan Shaw, David Sinn, and their colleagues in the Network Automation Forum are tackling this problem. Their goal is to... Read more »

One dark day, Ivan Pepelnjak stopped labbing. He just couldn’t make himself yet again go through assigning addresses, building links, putting devices in place, setting up OSPF, BGP, VXLAN, EVPN, etc. before even being able to start whatever simulation or test he wanted to do. But from that darkness arose netlab. Ivan created netlab to... Read more »

The days of network cowboy heroism are over… or at least they need to be. It’s time for network engineering to grow up and standardize how networks are built. Not only will this make life easier for all of us as we inherit networks when we move from company to company, but it’s the only... Read more »

Yale’s efforts to load-balance RADIUS servers is a case study in system design for resiliency. First, there was a lone, redundant PSN. Next, F5s load balancers entered the picture. Then the network team realized a feature in IOS-XE was the answer… and brought Cisco along the learning journey with them. Hear it all from the... Read more »

Guest Dinesh Dutt introduces his newest creation, SuzieQ. It’s a network observability platform application that has both a free, open source version and an enterprise version. Lightweight, fast, and platform-agnostic, SuzieQ’s use cases include network documentation, troubleshooting, fabric-wide visibility, network refresh and redesign, low/no code validation, audits and compliance, and proactive health checks. Hosts Ethan... Read more »

Remote and hybrid work means network engineers have to grapple with lossy residential networks such as home wireless that your work-from-home folks are using to access company resources. Their Wi-Fi sucks, and so their use of corporate resources sucks. Sure, you’ve got them plumbed into a SASE fabric, but that doesn’t fix their user experience... Read more »

On today’s episode, we discuss networking sources of truth. That’s right, sources of truth, because you’re likely to have more than one depending on your environment and your point of view. On LinkedIn, Ethan Banks quoted someone at the AutoCon0 conference who essentially said that the network itself shouldn’t be used as a source of... Read more »

At AutoCon0 in November 2023, guest Jeremy Schulman delivered a talk from the main stage about delivering network assurance. If the term “network assurance” doesn’t mean anything to you, think about how you prove after an install or a change that the network is doing what it’s supposed to be doing. If you’re doing it... Read more »

At NewYork-Presbyterian Hospital, patients are the priority. That focus on patient care extends to the hospital’s campus network, data center, wireless network, and SD-WAN. These networks are instrumental for delivering medical applications and connecting medical devices. On today’s Heavy Networking, we talk with network architects and engineers at NewYork-Presbyterian about their use of automation to... Read more »

SD-WAN is evolving to encompass more features and capabilities around security, application performance, network visibility, and more. On today’s Heavy Networking, sponsored by Palo Alto Networks, we look at how SD-WAN has transformed from a simple network connectivity solution to a comprehensive networking and security system. We discuss the limitations of legacy branch routers and... Read more »

Welcome to Heavy Networking! On today’s show we’ve got a roundtable conversation on the state of automation in the networking industry. This show was inspired by the recent AutoCon conference, which is a new conference focused specifically on network automation. Ethan Banks and I both attended, as did two our guests, and we’re going share... Read more »

Today we’re talking security, but security you don’t always see. Fortinet, today’s sponsor, has millions of devices in the field. These are real-world devices seeing real-world traffic, all day, everyday. While those devices have a primary protection role, they can also serve as sensors that collect threat signals and feed threat intelligence services that can,... Read more »

Public clouds abstract away much of the nitty-gritty work that goes into provisioning infrastructure, including networking. Application teams can quickly connect resources and deploy applications without having to know much about the plumbing that links everything together. When they compare the public cloud experience to standing up applications in an on-prem data center, the on-prem... Read more »

We continue our Future of Networking series with part two of our conversation with Brad Casemore. Now retired, Brad has participated in the industry as both a technologist and IDC analyst. In this episode we look at the rise of zero trust and what it means for campus networking. We discuss how geopolitics influences IT... Read more »

Cloud environments often have poor visibility and monitoring, and sensitive corporate data can be placed in many different locations – object stores, databases, and so on. Maintaining access control is hard, and managing compliance and governance on that access and data is a real problem. And each cloud has its own proprietary security tools that... Read more »

The Future of Networking series continues with Brad Casemore, who survived multiple decades in the technology sector, including sixteen years as an analyst for IDC. He’s been a longtime observer of networking markets, technologies, and trends. Recently retired from the analyst business, we’ve invited Brad onto the show as a kind of exit interview to... Read more »

AI and machine learning are finally being applied to networking in meaningful ways. On today's sponsored show we talk with Selector about its AIOps platform, which ingests networking logs, flows, configurations, SNMP, and other telemetry to detect patterns, spot problems, and provide contextual insights to help network engineers do their jobs better. We speak with Selector about what it is, how it works, and concrete use cases.

Today’s Heavy Networking is about collaborative automation via GitNops, which applies DevOps principles to networking. That means things like version control, working with sources of truth, operating infrastructure as code, and collaborating on network on changes and updates. GitNops benefits include automation, repeatability, and scalability. We'll dive into GitNops with guest Tom McGonagle.

Today on Heavy Networking we talk about how to get from a handful of Python scripts to a network automation platform that enables a self-service environment and incorporates the checks and governance required to make sure the automation doesn't blow up the network. Our sponsor today is Itential. Itential offers just such a network automation platform. We talk about how Itential makes it happen, including simple and more complex examples.

Today's Heavy Networking is another roundtable episode. We've assembled a group of network engineers to talk about what's on their minds. Topics today include why other IT departments adn end users are quick to blame the network first, and what can be done about it; using Nokia's open-source Containerlab for testing and development work; and why you shouldn't feel left behind when you hear talk about 400G and 800G networks.

Today’s Heavy Networking, sponsored by Palo Alto Networks, discusses Zero Trust Network Access (ZTNA) across the Secure Access Service Edge (SASE). You could think of ZTNA, as VPN evolved and SASE as SD-WAN evolved. Not only do you get an overlay connectivity fabric, but you also get cloud-hosted security services. We talk about how it all works together, the role of client software, Palo Alto Networks' architecture, integrating an IDP, and more.

Today's Heavy Networking covers Pandas. Not the cuddly bears that eat bamboo, but the Python library that makes it easy for you to work with a set of data. Import Pandas at the top of your Python script, follow one of many Pandas tutorials online, and in short order you’ll be able to perform data operations in a spreadsheet-like way. We talk network automation use cases for Pandas with Rick Donato.

Our topic today on Heavy Networking is SD-WAN monitoring at massive scale. Scale can grow quickly with SD-WAN when you account for the underlay, overlays, gateways, endpoints, and more. We talk with sponsor Broadcom about their monitoring platform and dig into a case study with a Broadcom customer providing global IT infrastructure for thousands of their own customers.

On today's sponsored Heavy Networking we talk with Fortinet about how it converges network and security across the breadth of its portfolio. Fortinet is best known for its firewalls, but the company also offers campus and branch switches, has a wireless portfolio, offers SD-WAN and SASE, provides AI-supported operations, and more. We also talk about Fortinet's approach to the branch office that rethinks networking, security, and performance.

Virtual Application Networks, or VANs, are today’s Heavy Networking topic. Our guest is Ted Ross, motive force behind the Skupper.io project. Skupper builds VANs in Kubernetes clusters that are conceptually like a VLAN or VPN, except that all the magic happens at layer 7. Skupper is based on the Advanced Message Queueing Protocol (AMQP), making it effectively a message bus used to interconnect application messages inside of mTLS tunnels running on top of whatever L3 network is available. If you're confused, don't be. We talk it all out, and explain why it's relevant to today's networking pros.

On today’s Heavy Networking we explore the edge. But where is the edge? In today's conversation with sponsor VMware, it's a remote location where data is being generated. It could be end users in a branch office, or IoT devices and sensors on a factory floor. These edge locations needs integrated compute and networking to run application workloads while also being able to  connect to cloud applications and services. We discuss the VMware Edge Cloud Orchestrator, which provides unified management for networking and compute at remote sites.

In today's sponsored Heavy Networking we explore new features in Cisco Thousand Eyes, an operational tool based on visibility and observability of public and private network. Thousand Eyes has continued to grow into complex operational areas such AWS Network Path, Webex performance, and integrations with Meraki to help you identify and fix network and application performance problems.

EVPN/VXLAN is our topic on today's Heavy Networking. What is it? What’s it for? Should you deploy it? Since you’ve probably already got a network, how do you add EVPN to it? Do you need special hardware? How does EVPN impact your security design? And what are the fundamentals? Our guest with the answers is IT instructor Tony Bourke.

On today's Heavy Networking we explore network and firewall automation with sponsor BackBox. BackBox has developed a platform that aims to deliver practical automation out of the box. We get under the hood to understand how it works, what it delivers, and how it addresses the challenges of network and security operations.

This week on Heavy Networking we've assembled a roundtable of network engineers to talk about...stuff. Each guest has brought a topic to discuss with the table, so we've got lots of subjects and lots of experiences and opinions. In particular we explore SPB, career advice, getting network automation off the ground, and the joys and perils of self-hosting.

On today's sponsored Heavy Networking we dig into cloud-delivered Secure Web Gateways (SWGs), which help guard end users against Web-based threats and enforce corporate Web access policies. As employees split time between home, office, and who knows where else, and as more applications move online, cloud-based SWGs help connect and protect workers. Our sponsor is Palo Alto Networks.

If you’ve been staring down the barrel of network automation and wonder what the proper approach might be, today’s episode is for you. The Packet Pushers chat with Tony Bourke about what network automation tools and techniques have become the default standard, how to prepare your network and team for automation, and how to get started.

On today’s Heavy Networking podcast, we look at how sponsor ZPE Systems is rethinking Out-Of-Band management for automated, NetOps-driven infrastructure. This includes tasks like device staging, deployments, upgrades, and more. And you don’t just have to take ZPE’s word for it; we also talk to a customer who’s using the products to run a retail business with a lean networking team that supports more than fifty sites.

On today’s Heavy Networking we talk LACP and link aggregation. While bonding two or more links together to act as a single virtual link has been done for decades, LACP and link aggregation aren't the same thing, and the distinction matters. Our guest to get into the differences is network instructor Tony Bourke.

Certifications are a part of life in IT. On today's Heavy Networking we explore preparation strategies with guest Mary Fasang. Her certs run the gamut from CompTIA to MCSE to the CCNP, as well as the PMP and ITIL certs. How should you prepare for a cert in 2023 when there’s so much content, so many training options, as well as home labbing available? How do you handle failure? Which certs have been the hardest? What study materials have proved helpful? Mary shares her strategies.

Packet-level fundamentals are essential for network engineers to be able to diagnose and solve network and application problems. On today's Heavy Networking, we dive into the transport layer and packets with packet analysis expert and instructor Chris Greer.

Its about reducing the cost and complexity of DWDM coherent optical networks. Connecting the DWDM network directly to your router removes the DWDM edge equipment which simplifies operation, reduce cost,space & power while improving provisioning time. How is Juniper entering this market and what do you need to know ?

Today we’re going deep on software-defined networking for containers and OpenStack with sponsor Juniper Networks. Juniper has revamped its approach to secure networking for telcos and telco cloud-delivered services with Juniper’s Cloud-Native Contrail Networking or CN2 software. CN2 lets you automate the creation of network connections for containers and for virtual machines while also providing routing, security, segmentation and isolation of workloads. Our guest and guide into the guts of Cloud-Native Contrail Networking, hereafter referred to as CN2, is Nick Davey. Nick is Director of Product Management for SDN and Telco Cloud technologies.

Remote operation of infrastructure has renewed importance in the era of remote working. Opengear offers secure, zero trust and segmented methods to reach serial & LAN ports plus GUI interfaces. You can add observability agents like Thousand Eyes into containers so that your worst day becomes just another day.

By some estimates, 50 to 70 million tons of e-waste is generated every year, and that number is growing. When sent to landfills to be buried or burned, e-waste can leach toxic chemicals into soil and air. On today’s Heavy Networking, we’ll look at options for responsible disposal of IT gear, including repurposing it on site, reselling or donating it, and working with e-cycling companies.

On today’s Heavy Networking we have a conversation about monitoring, visibility, and observability with sponsor Palo Alto Networks. More specifically we’ll dig into Palo Alto Networks’ Autonomous Digital Experience Management, or ADEM product, and how Palo Alto Networks is integrating ADEM with AIOps.

If you’ve shied away from network automation because you’re a network engineer not a coder, fear not. There are network automation approaches that can help you get needful work done and don’t require a computer science degree. On today's Heavy Networking we talk with sponsor BackBox about its practical approach to network automation and dive into use cases including network OS backups and compliance.

Modern F1 racing is bathed in data. Real-time vehicle telemetry. In-car video feeds. More video from camera crews and drones. Live streaming. All of this and more needs a network. On today's Heavy Networking we speak with Formula 1's David Ramsden, Senior Network Engineer; and Lee Wright, Head of IT Operations, to get the inside track on building the networks that support F1 races.

It's common for SD-WAN vendors to offer monitoring as part of the solution, but leaves the question … how do I monitor the rest of the network? Today’s sponsor Broadcom offers digital experience monitoring that is independent of the underlying WAN infrastructure. We explore how it works with guest is Jeremy Rossbach, Chief Technical Evangelist, NetOps by Broadcom.

Today's Heavy Networking explores the challenges Wireless ISPs (WISPs) face when bringing services to locations without a lot of fiber in the ground. Those challenges include lining up radio antennas, maintaining mountaintop gear, wild horses, network and routing designs, and more. Our guest is Bradley Thompson, Senior Network Engineer at SkyFiber Internet.

On today's Heavy Networking we dive into the frameworks commonly used by service providers to tackle network slicing and traffic engineering challenges. We'll also talk their pros and cons, and the approach that Cisco is seeing its customers adopt as providers create virtual networking products for their customers. Cisco is our sponsor for today's show.

You ever want a group of fellow networking nerds to hang with once in a while? The US Networking User Association might be exactly what you’re looking for. With local networking user groups popping up in various places all over the US and soon other countries, the USNUA is fostering community and knowledge sharing for networkers everywhere. On today's Heavy Networking we speak with Jason Gintert and Chris Kane, two of the folks behind the USNUA organization, to discuss what the USNUA is, and how you can work with them to get a NUG started in your area.

Fortinet is a security vendor most of you have heard of. But if all you think of when you hear the name “Fortinet” is firewalls, well yeah, but you should think more broadly. On today's sponsored Heavy Networking we're going after the work-from-anywhere challenge with Fortinet’s Zero Trust Network Access (ZTNA) and Secure Access Service Edge (SASE) tech, all of which is baked into FortiOS. If you’re running a FortiGate, you’ve got these capabilities already.

Today’s Heavy Networking podcast explores the concept of NetDevOps with sponsor Itential. The idea behind NetDevOps is to advance your network to the point where it’s self-service; that is, the network you operate can be consumed the way public cloud services are consumed. Our guest is Itential co-founder & CTO Chris Wade.

On today’s Heavy Networking we get into IPv6 essentials for network engineers, including how to incorporate IPv6 support in upcoming projects, how IPv6 affects NAT and subnetting, what the heck Happy Eyeballs and nibble boundaries are, and why you should approach IPv6 with a mindset of abundance not scarcity.

You know all those Apple and other IoT devices connected to your wireless network? Lots of them run apps that discover services on your network via multicast DNS (mDNS). All of that mDNS traffic can have a significant impact on your WLAN’s performance. On today's Heavy Networking we talk with guest Bryan Ward who has actually measured the impact of mDNS on a production wireless network to see what would happen if he let mDNS traffic run wild.

Lots of folks suffer from impostor syndrome. Tech is complex--how could you know what you’re doing? And yet, many of us are responsible for incredibly complex IT systems. Fake it ‘til you make it, right? To handle the cognitive dissonance of impostor syndrome, we overcompensate. In doing so, we pay a personal price. Today's Heavy Networking guest is Matt Vitale. He's here to share what he's learned about coping with and overcoming imposter syndrome.

On today's Heavy Networking we sub in a podcast from our Heavy Strategy channel. Greg Ferro from the Packet Pushers and Johna Till Johnson, CEO of Nemertes Research, discuss the impacts of ChatGPT and AI on the technology workspace, including whether human workers can partner with these tools to increase productivity and improve technology experiences. They don't have answers, but they do have unanswered questions.

On today’s Heavy Networking I talk with two people who have built and are running Hachyderm, a Mastodon instance which orients itself towards technical-minded folks. What started as a cool project in the basement suddenly grew to tens of thousands of users. I talk with Hachyderm's founders on how they scaled, problems they encountered, how they solved problems, and how they grappled with technical and human challenges.

On today's Heavy Networking, sponsor Graphaint is here to discuss how it's rethinking SD-WAN using a Network-as-a-Service (NaaS) model to help customers consume WAN connectivity instead of setting up and operating a WAN for themselves. Graphiant combines SD-WAN mainstays such as multiple connection options and traffic policies based on applications and performance, without the operational overhead of SD-WAN tunnels. We dive into the how and why in this episode.

On Heavy Networking today we look at a home lab running VMware products including NSX, as well as infrastructure-as-code products Terraform, Packer, and Ansible. These use cases create a different hardware demand than virtualized network operating system images. Guest Maarten Van Driessen explains it all, including how he saves money on lab gear.

In today’s Heavy Networking show with sponsor Broadcom we go deep into network management and Digital Experience Monitoring (DEM). At its heart, DEM is about understanding the user experience of the network. Network monitoring and management products that incorporate user experience, such as Broadcom's DX NetOps, can provide visibility into network and application performance to help inform troubleshooting efforts and speed resolution.

Welcome to Heavy Networking! In this episode we discuss LibreQoS, a free and open source software project to help ISPs improve network latency and responsiveness and improve the Quality of Experience (QoE) for their customers. That project is LibreQoS, and it's being used by ISPs and others to ensure stable latency across networks.

The Packet Pushers' Heavy Networking podcast dives into sponsor Augtera and how its AI platform, purpose-built for networking, improves network operations and enables automation. We'll examine how Augtera works, how it aims to move beyond the automation of configurations to automate operations and fault management, the kinds of data it collects and how, and how customers are using Augtera in production networks

Today's Heavy Networking is a forward-looking episode about semantic networking. Semantic networking aims to make decisions on how to route packets based on more than just the destination address and give network operators more routing choices based on considerations such as bandwidth, cost, performance, application type, and so on. But how do you add semantic information to IP headers? How do you program routers and networking hardware to consume semantics? Do we really need this? Guests Adrian Farrel and Hannes Gredler join Greg Ferro and Ethan Banks to discuss and debate.

Would you give an artificial intelligence responsibility to write your router configurations? You wouldn’t. Not yet. But we’re not as far from that as you might think. On today's Heavy Networking we dig into OpenAI and ChatGPT and what it might mean for networking.

On today's Heavy Networking, guest Tom Costello walks us through his personal lab in the cloud. We discuss pros and cons of the cloud vs. a basement, using EVE-NG vs. CML, how automation can help you manage cloud costs, resources for those who might want a cloud lab, and more.

Today on Heavy Networking, home labs. Specifically, we’re going to discuss building and using a lab with virtualized Juniper gear in EVE-NG. Our guest is Christian Scholz, who has written and presented quite a bit on this topic.

On today’s Heavy Networking we get inside the process of technical writing and product documentation from a person deep in the trenches of creating docs. We also talk about why writing as a skill might be worth cultivating even if you aren’t responsible for creating the manual.

On today's Heavy Networking podcast, sponsored by Juniper Networks, we discuss how metro networks are evolving to Cloud Metro, how to apply cloud principles and automation to these locations, and how the attack surface changes when you transition to Cloud Metro. We also talk about how Juniper supports active assurance and zero trust security in these environments.

Today on Heavy Networking, modeling your network with Batfish. Batfish is an open-source project that builds a model of your production network based on a snapshot of your network and device configurations. Engineers can then test and validate changes against this model before making those changes to the production network. Guest Rick Donato walks us through how Batfish works, use cases, and more.

Today on Heavy Networking, a discussion with sponsor VMware about SD-WAN and SASE. We’re diving into announcements from VMware Explore Barcelona 2022 covering a new SD-WAN client and more. With this client, you’ll be able to connect your users to the SASE cloud with software--no hardware edge box required. We dive into how it works, the network architecture, use cases, and more.

On today's Heavy Networking we look at the idea of embedding zero trust into applications. The way we do cyber security these days has failed in significant ways. What if we could extend the AAA or RBAC model to all applications? Better yet, what if we take the RBAC model, make authentication more robust than username & password, assess endpoint security posture constantly, and evaluate each request individually up at layer 7 for all applications? Guest Galeal Zino has opinions on what embedded zero trust looks like. We discuss.

Heavy Networking welcomes sponsor Netris to the show with a special episode for you network nerds who are really getting into automation, infrastructure as code, pipelines, and so on. Netris is all about bringing that public cloud VPC experience to the network you’ve already got. Imagine being able to consume your existing network with APIs and being able to stand up VLANs, VXLANs, elastic load balancers, firewalls, Internet gateways, and more the same way you do in the cloud, but on premises.

Today on Heavy Networking we have a round-table discussion about Data Processing Units (DPUs). These devices let you bring networking, security, and storage capabilities closer to server workloads without burdening server CPUs. Guests Pete Lumbis and Aaron Glenn help us dissect DPUs, identify use cases, discuss adoption and operational challenges, and more.

Today’s Heavy Networking, sponsored by Nokia, dives into Nokia's fabric-based approach to data center automation and operations. That approach includes its SR Linux network OS, its Fabric Services System intent-based platform, its NetOps Development Kit, or NDK, and how all this ties together to address your operational life cycle across Day zero, Day 1, Day Two, and beyond.

On today’s Heavy Networking episode, I talk with Nick Carter about Flock Networks, his routing protocol stack startup, as well as Nick’s love of the Rust programming language. As a network engineer, maybe you don’t think you care about Rust. Nick’s here to explain why the discerning network engineer might prefer their routing daemons to have been written in Rust. We also talk about the pleasures and travails of startup life.

As carriers and service providers look to embrace disaggregated infrastructure and software, and drive new business through technologies such as network slicing, it’s critical to have management and orchestration capabilities to coordinate  hardware and software resources in the RAN, the transport layer, and the network core. On today's sponsored Heavy Networking, sponsored by Juniper, we dive into Juniper's Service Management and Orchestration (SMO) platform, which is designed to provision, manage and monetize custom network services on demand.

On today’s Heavy Networking podcast, Kevin Myers joins us for a whitebox conversation. Kevin helps Internet Service Providers build their networks, and has noticed increased adoption of whitebox switches. Why? Are the problems whitebox solves for these ISPs the same you might have at your company? Should you consider whitebox instead of Cisco, Juniper, or Arista? Maybe…and maybe not.

On today’s sponsored Heavy Networking we dive into new features of Cisco’s Network Services Orchestrator, or NSO. NSO is a network orchestration platform for large enterprises and service providers that offers multi-vendor automation, configuration management, service creation, ongoing monitoring and upgrades, and more.