0

Comparing EVPN with Flood-and-Learn Fabrics

One of ipSpace.net subscribers sent me this question after watching the EVPN Technical Deep Dive webinar:

Do you have a writeup that compares and contrasts the hardware resource utilization when one uses flood-and-learn or BGP EVPN in a leaf-and-spine network?

I don’t… so let’s fix that omission. In this blog post we’ll focus on pure layer-2 forwarding (aka bridging), a follow-up blog post will describe the implications of adding EVPN IP functionality.

0

When we’ll get a 128-bit CPU

On Hacker News, this article claiming "You won't live to see a 128-bit CPU" is trending". Sadly, it was non-technical, so didn't really contain anything useful. I thought I'd write up some technical notes.

The issue isn't the CPU, but memory. It's not about the size of computations, but when CPUs will need more than 64-bits to address all the memory future computers will have. It's a simple question of math and Moore's Law.

Today, Intel's server CPUs support 48-bit addresses, which is enough to address 256-terabytes of memory -- in theory. In practice, Amazon's AWS cloud servers are offered up to 24-terabytes, or 45-bit addresses, in the year 2020.

Doing the math, it means we have 19-bits or 38-years left before we exceed the 64-bit registers in modern processors. This means that by the year 2058, we'll exceed the current address size and need to move 128-bits. Most people reading this blogpost will be alive to see that, though probably retired.

There are lots of reasons to suspect that this event will come both sooner and later.

It could come sooner if storage merges with memory. We are moving away from rotating platters of rust toward solid-state Continue reading

0

Worth Reading: Machine Learning Deserves Better Than This

This article is totally unrelated to networking, and describes how medical researchers misuse machine learning hype to publish two-column snake oil. Any correlation with AI/ML in networking is purely coincidental.

Machine Learning Deserves Better Than This

0

Worth Reading: Machine Learning Deserves Better Than This

This article is totally unrelated to networking, and describes how medical researchers misuse machine learning hype to publish two-column snake oil. Any correlation with AI/ML in networking is purely coincidental.

Machine Learning Deserves Better Than This

0

Worth Reading: Is Your Consultant a Parasite?

Stumbled upon a must-read article: Is Your Consultant a Parasite?

For an even more snarky take on the subject, enjoy the Ten basic rules for dealing with strategy consultants by Simon Wardley.

0

Worth Reading: Is Your Consultant a Parasite?

Stumbled upon a must-read article: Is Your Consultant a Parasite?

For an even more snarky take on the subject, enjoy the Ten basic rules for dealing with strategy consultants by Simon Wardley.

0

SSH Algorithm and Cipher Options

Have you ever tried to SSH to an network device and recieved the dreaded Unable to negotiate with <user> port 22: no matching key exchange method found. Their offer: <key-algorithm>. In this post ill cover how to work around this issue. Key Algorithms Specify the Key...continue reading

0

Lenovo Gives SuperMUC-NG Supercomputer An Upgrade

Lenovo became a powerhouse in the HPC and supercomputer spaces back in 2014, when it bought IBM’s System x server division for $2.1 billion in a deal that also saw it license storage and system management software from Big Blue. …

Lenovo Gives SuperMUC-NG Supercomputer An Upgrade was written by Jeffrey Burt at The Next Platform.

0

Think Locally About Data Management, But Act Globally

The amount of data being created is continuing to rise exponentially and it’s coming in all shapes and sizes and from myriad locations. …

Think Locally About Data Management, But Act Globally was written by Jeffrey Burt at The Next Platform.

0

Observe & Troubleshoot Your Kubernetes Environments with Dynamic Service Graph

Kubernetes workloads are highly dynamic, ephemeral, and are deployed on a distributed and agile infrastructure. Application developers, DevOps teams, and site reliability engineers (SREs) often require better visibility of their different microservices, what their dependencies are, how they are interconnected, and which other clients and applications access them. This makes Kubernetes observability challenges unique. While Kubernetes helps to meet the needs of deploying and managing distributed applications, its observability challenges require a Kubernetes-native approach.

Traditional monitoring and observability solutions create data silos by collecting data at different levels (e.g. infrastructure, cluster, and application levels), or from a large number of ephemeral objects that generate data across a distributed environment. Traditional monitoring and observability solutions then stitch this data together to provide a near real-time snapshot view. This approach is not scalable given the high volume of granular data generated at each level, as well as Kubernetes’ distributed nature. It also starts to become expensive and budget unfriendly to run traditional monitoring solutions, as they require higher resource consumption (high-performance memory, more compute, and higher bandwidth).

In contrast, a Kubernetes-native observability solution can visualize all information with all relationship context intact and provide a high-fidelity view of the environment. This Continue reading

0

Weekend Reads 061821

Electric vehicles are expected to account for 58% of global passenger vehicle sales by 2040. The software and electrical components markets are also likely to face increased pressure and new challenges as they develop secure designs and equipment for these futuristic vehicles.

Community managers, maintainers, and foundations seek metrics and insights about open source communities.

The principle of least privilege in cybersecurity prescribes that no user should have access to system resources beyond what’s necessary for fulfilling a specific task.

(External) memory fragmentation is a long-standing Linux kernel programming issue. As the system runs, it assigns various tasks to memory pages.

In a world that is constantly evaluating costs, it is little wonder that there is an increasing demand for cost-effective solutions to business problems. In the real world, this means ‘free,’ and in the digital marketplace, it means ‘open source.’

There’s an infinite number of studies of ransomware lately, all breathlessly talking about how to fight this dangerous threat. They’re all dangerously wrong. Ransomware is not the problem.

Researchers have disclosed a new type of attack that exploits misconfigurations in transport layer security (TLS) servers to redirect HTTPS traffic from a victim’s web browser to a different TLS Continue reading

0

Heavy Networking 583: How Salesforce Evolved Its Branch Network With Prisma SD-WAN (Sponsored)

On today's Heavy Networking, sponsored by Palo Alto Networks, we hear from Salesforce about the evolution of its branch network to SD-WAN. Salesforce was able to trade MPLS for Internet broadband, get more bandwidth for less money, employ application-based steering and policy enforcement, and more. Our guests are Georgi Stoev, Sr. Network Architect at Salesforce; and Kumar Ramachandran, Senior Vice President at Palo Alto Networks.

0

Heavy Networking 583: How Salesforce Evolved Its Branch Network With Prisma SD-WAN (Sponsored)

On today's Heavy Networking, sponsored by Palo Alto Networks, we hear from Salesforce about the evolution of its branch network to SD-WAN. Salesforce was able to trade MPLS for Internet broadband, get more bandwidth for less money, employ application-based steering and policy enforcement, and more. Our guests are Georgi Stoev, Sr. Network Architect at Salesforce; and Kumar Ramachandran, Senior Vice President at Palo Alto Networks.

The post Heavy Networking 583: How Salesforce Evolved Its Branch Network With Prisma SD-WAN (Sponsored) appeared first on Packet Pushers.

0

Don’t OutSMART Your Goals

I read a piece on LifeHacker yesterday that made me shake my head a bit. I’m sure the title SMART Goals Are Overrated was designed to get people to click on it, so from that perspective it succeeded. Wading into the discourse there was an outline of how SMART goals were originally designed for managers to give tasks to employees and how SMART doesn’t fit every goal you might want to set, especially personal aspirational ones. Since I have a lot of experience with using SMART goals both for myself and for others I wanted to give some perspective on why SMART may not be the best way to go for everything but you’re a fool if you don’t at least use it as a measuring tool.

SMRT, Eh?

As a recap, SMART is an acronym for the five key things you need to apply to your goal:

• S – Specific (what are you going to do)
• M – Measurable (how will you know when you’ve succeeded)
• A – Attainable or Assignable (can you or the person you’ve selected do this thing)
• R – Relevant or Relatable (is this goal appropriate for me or for the person doing it)
• T Continue reading

0

Observe & Troubleshoot Your Kubernetes Environments with Dynamic Service Graph

Kubernetes workloads are highly dynamic, ephemeral, and are deployed on a distributed and agile infrastructure. Application developers, DevOps teams, and site reliability engineers (SREs) often require better visibility of their different microservices, what their dependencies are, how they are interconnected, and which other clients and applications access them. This makes Kubernetes observability challenges unique. While Kubernetes helps to meet the needs of deploying and managing distributed applications, its observability challenges require a Kubernetes-native approach.

Traditional monitoring and observability solutions create data silos by collecting data at different levels (e.g. infrastructure, cluster, and application levels), or from a large number of ephemeral objects that generate data across a distributed environment. Traditional monitoring and observability solutions then stitch this data together to provide a near real-time snapshot view. This approach is not scalable given the high volume of granular data generated at each level, as well as Kubernetes’ distributed nature. It also starts to become expensive and budget unfriendly to run traditional monitoring solutions, as they require higher resource consumption (high-performance memory, more compute, and higher bandwidth).

In contrast, a Kubernetes-native observability solution can visualize all information with all relationship context intact and provide a high-fidelity view of the environment. This Continue reading

0

EPCC Kicking Tires on New CS-1 AI System

The Edinburgh Parallel Computing Center (EPPC) is up and running with its Cerebras CS-1 waferscale system and is already working with European companies in biomedical and cybersecurity arenas in addition to its own research into different programming and AI models and projects in natural language processing and genome-wide association studies. …

EPCC Kicking Tires on New CS-1 AI System was written by Nicole Hemsoth at The Next Platform.

0

Modern Network Automation is Missing from Modern IT

Delivering automation throughout the IT stack helps alleviate organizational strains, making it possible to manage change at scale both on-premises and in the cloud while ensuring the performance and functionality.

0

Technology Short Take 141

Welcome to Technology Short Take #141! This is the first Technology Short Take compiled, written, and published entirely on my M1-based MacBook Pro (see my review here). The collection of links shared below covers a fairly wide range of topics, from old Sun hardware to working with serverless frameworks in the public cloud. I hope that you find something useful here. Enjoy!

Networking

• Via Ivan Pepelnjak, I was pointed to Jon Langemak’s in-depth discussion of working with Linux VRFs.
• I read a couple of Cilium-related blog posts recently that may be useful. The first is a post on Cilium and F5 load balancer integration, while the second discusses implementing Kubernetes network policies with Cilium and Linkerd.
• Sonia Cuff provides a set of links for detailed instructions on setting up VPN access from macOS to Microsoft Azure with Azure Active Directory authentication.
• Patrick Kremer walks through using Postman to implement BGP route filtering with VMware Cloud on AWS. I think this post is interesting/useful for a couple different reasons. One, the content is useful in and of itself. Second, the content reflects—in my opinion—the changing nature of what it means to be a “networking professional.” It’s not just Continue reading

0

Cloudflare TV: Doing it Live, 1,000 Times and Counting

Last week, Cloudflare TV celebrated its first anniversary the only way it knows how: with a broadcast brimming with live programming spanning everything from the keynotes of Cloudflare Connect, to a day-long virtual career fair, to our flagship game show Silicon Valley Squares.

When our co-founder and CEO Matthew Prince introduced Cloudflare TV to the world last year, he described it as a platform for experimentation. By empowering Cloudflare employees to try whatever they could think up on air — bound only by restraints of common sense — we hoped to unlock aspects of our team’s talent and creativity that otherwise might go untapped in the midst of the pandemic.

The results, as they say, have been extraordinary.

Since launching in June 2020, Cloudflare TV has featured over 1,000 original live episodes covering an incredible array of topics: technical deep dives and tutorials like Hardware at Cloudflare, Leveling up Web Performance with HTTP/3, and Hacker Time. Security expertise from top CISOs and compliance experts. In-depth policy discussions. And of course, updates on Cloudflare’s products with weekly episodes of Latest from Product and Engineering, Estas Semanas en Cloudflare en Español, and launch-day introductions to Magic WAN Continue reading

0

Member News: Somalia Chapter Focuses on Internet Education

Paying by phone: The Somalia chapter of the Internet Society is focusing on educating Internet users, particularly young people, about mobile payments. The chapter is working with technology stakeholders to improve digital literacy. The chapter notes that 70 percent of adults in Somalia use mobile money services regularly, and more than two thirds of payments […]

The post Member News: Somalia Chapter Focuses on Internet Education appeared first on Internet Society.