Preloading Extra Images with Kubernetes Image Builder

The Image Builder project is a set of tools aimed at automating the creation of Kubernetes disk images—such as VM templates or Amazon Machine Images (AMIs). (Interesting side note: Image Builder is the evolution of a much older Heptio project where I was a minor contributor.) I recently had a need to build a custom AMI with some extra container images preloaded, and in this post I’ll share with you how to configure Image Builder to preload additional container images.

Image Builder isn’t a single binary; it’s a framework built on top of other tools such as Packer and Ansible. Although in this post I’m discussing Image Builder in the context of building an AMI, it’s not limited to use with AWS. You can use Image Builder for a pretty wide collection of platforms (check the Image Builder web site for more details).

To have Image Builder preload additional images into your disk image, there are three changes needed. All three of these changes belong in the images/capi/packer/config/additional_components.json file:

  1. Set load_additional_components to true. (The default value is false.)
  2. Set additional_registry_images to true. (This also defaults to false.)
  3. Set additional_registry_images_list to a comma-delimited list of fully-qualified image Continue reading

Common Services VRF with EVPN Control Plane

After discovering that some EVPN implementations support multiple transit VNI values in a single VRF, I had to check whether I could implement a common services L3VPN with EVPN.

A common services VPN is a VPN in which server sites can communicate with each other and the clients, but the clients cannot communicate between themselves.

TL&DR: It works (on Arista cEOS)1.

Here are the relevant parts of a netlab lab topology I used in my test (you can find the complete lab topology in netlab-examples GitHub repository):

NAN072: Exploring Internet2’s Network Automation for University and Research Networks

On today’s episode we delve into Internet2’s mission to advance research and education through high-speed connectivity. University and research networks may have more complicated designs, requirements, and restrictions than you’d expect. We talk about these challenges, along with innovative network automation solutions. The discussion emphasizes the collaborative, member-driven approach of Internet2 and its commitment to... Read more »

Not All OSPF Inter-area Traffic Traverses Interfaces In Area 0

Everyone knows that OSPF is a link state protocol. Those that study also discover that OSPF behaves like distance vector between areas as Type-1- and Type-2 LSAs are not flooded between areas, but rather summarized in Type-3 LSAs. This means that OSPF is a logical star, or hub with spokes, where Area 0 is the backbone and all other areas must connect to Area 0. This is shown below:

With this topology, since all the areas only connect to the backbone area, traffic between areas must traverse the backbone:

We learn about this behavior in literature where there is a very straight forward topology where each ABR is only attached to one area beyond the backbone. Such a topology is shown below:

In such a topology, traffic between RT04 and RT05 has to traverse the backbone. This is shown below:

However, what if you have a topology which is not as clear cut? Where an ABR attaches to multiple areas? This is what we will explore in this post. We’ll be using the topology below:

In this topology, RT02 and RT03 are ABRs. RT02 is attached to both Area 1 and Area 2 in addition to the backbone, while RT03 Continue reading

Nvidia Rolls Out Blueprints For The Next Wave Of Generative AI

Hardware is always the star of Nvidia’s GPU Technology Conference, and this year we got previews of “Blackwell” datacenter GPUs, the cornerstone of a 2025 platform that includes “Grace” CPUs, the NVLink Switch 5 chip, the Bluefield-3 DPU, and other components, all of which Nvidia is talking about again this week at the Hot Chips 2024 conference.

Nvidia Rolls Out Blueprints For The Next Wave Of Generative AI was written by Jeffrey Burt at The Next Platform.

VMware Private Cloud Now Has a Catalog of Advanced Services

Tired of dealing with cloud providers and mulling a move to a private cloud instead? Broadcom wants you to take a look at its operation of a private cloud. This week at Paul Turner, Broadcom vice president of products for VCF, in a press briefing. Broadcom is positioning VCF as a lower-cost, more secure alternative to public cloud computing. Overall, the goal is to help the organization create an infrastructure that works together as a single, unified whole while supporting modern application architectures. Virtual Cloud Foundation architecture (VMware) Big Results Moving to a Private Cloud According to the company, a private cloud approach can result in: Continue reading

Multivendor EVPN Just Works

Shipping netlab release 1.9.0 included running 36 hours of integration tests, including fifteen VXLAN/EVPN tests covering:

  • Bridging multiple VLANs
  • Asymmetric IRB, symmetric IRB, central routing, and running OSPF within an IRB VRF.
  • Layer-3 only VPN, including routing protocols (OSPF and BGP) between PE-router and CE-routers
  • All designs evangelized by the vendors: IBGP+OSPF, EBGP-only (including reusing BGP AS number on leaves), EBGP over the interface (unnumbered) BGP sessions, IBGP-over-EBGP, and EBGP-over-EBGP.

All tests included one or two devices under test and one or more FRR containers1 running EVPN/VXLAN with the devices under test. The results were phenomenal; apart from a few exceptions, everything Just Worked™️.

NB492: AMD Spends Big For Engineering Talent; FCC Cracks Down on AI-Faked Robocalls

Take a Network Break! A US appeals court will let a privacy lawsuit against Google go forward, striking telecom workers are a reminder to test your resiliency planning, and AMD spends nearly $5 billion to acquire systems engineering talent from ZT Systems. Juniper Networks offers deep discounts and other enticements to get network engineers to... Read more »