Technology Short Take 179

Welcome to Technology Short Take #179! I’m back with another set of links to articles on various data center- and IT-related topics. In the interest of full transparency, I’d like to give credit to Russ White for his “Weekend Reads” series of posts, which are similar in nature to my Technology Short Takes. If you aren’t reading Russ’ “Weekend Reads” posts, you’re missing out on a good source of useful information. Several of the links included below are taken from recent posts by Russ. Thanks, Russ—and to all the other content creators and content curators referenced here—for your great work! Now, on to the content.

Networking

  • This post about netlab just reminds me that I really should spend some quality time with it. Need more than 24 hours in a day…
  • Timothy Ham created a GitHub Gist-based short IPv6 guide for home IPv4 admins.
  • Daniel Dib has a six-part series (so far) on Cisco vPC in a VXLAN/EVPN network. I’ll only link to part 1 of the series; you can find links to the rest in the sidebar on his site. I haven’t yet read all of them, but they’re on my list to read.
  • Hat tip to Ivan Continue reading

How the first 2024 US presidential debate influenced Internet traffic and security trends

Key findings:

  • The Biden vs. Trump debate influenced Internet traffic at the state level in the US, with drops in traffic as high as 17% (in Vermont) during the debate.
  • Microblogging and video streaming platforms saw traffic changes during the debate.
  • Trump-related sites, including donation platforms, gained much more traction than Biden’s during and after the debate.
  • Emails with “Trump” in the subject had higher rates of spam and malicious content compared to those with “Biden.”
  • No increase in cyberattacks during the debate, but frequent DDoS attacks targeted government and political sites in the preceding months.

Internet traffic ebbs and flows usually follow human patterns, and high visibility events that are broadcast on TV usually have an impact. Let’s take a look at the first of the 2024 United States presidential debates between the two major presumptive candidates, Joe Biden and Donald Trump, for the November presidential election.

2024 has been dubbed “the year of elections,” with elections taking place in over 60 countries, as we have mentioned before (1, 2, 3). We are regularly updating our election report on Cloudflare Radar, including analysis of recent elections in South Africa, India, Iceland, Mexico, Continue reading

Supporting Postgres Named Prepared Statements in Hyperdrive

Hyperdrive (Cloudflare’s globally distributed SQL connection pooler and cache) recently added support for Postgres protocol-level named prepared statements across pooled connections. Named prepared statements allow Postgres to cache query execution plans, providing potentially substantial performance improvements. Further, many popular drivers in the ecosystem use these by default, meaning that not having them is a bit of a footgun for developers. We are very excited that Hyperdrive’s users will now have access to better performance and a more seamless development experience, without needing to make any significant changes to their applications!

While we're not the first connection pooler to add this support (PgBouncer got to it in October 2023 in version 1.21, for example), there were some unique challenges in how we implemented it. To that end, we wanted to do a deep dive on what it took for us to deliver this.

Hyper-what?

One of the classic problems of building on the web is that your users are everywhere, but your database tends to be in one spot.  Combine that with pesky limitations like network routing, or the speed of light, and you can often run into situations where your users feel the pain of having your Continue reading

Embedded function calling in Workers AI: easier, smarter, faster

Introducing embedded function calling and a new ai-utils package

Today, we’re excited to announce a novel way to do function calling that co-locates LLM inference with function execution, and a new ai-utils package that upgrades the developer experience for function calling.

This is a follow-up to our mid-June announcement for traditional function calling, which allows you to leverage a Large Language Model (LLM) to intelligently generate structured outputs and pass them to an API call. Function calling has been largely adopted and standardized in the industry as a way for AI models to help perform actions on behalf of a user.

Our goal is to make building with AI as easy as possible, which is why we’re introducing a new @cloudflare/ai-utils npm package that allows developers to get started quickly with embedded function calling. These helper tools drastically simplify your workflow by actually executing your function code and dynamically generating tools from OpenAPI specs. We’ve also open-sourced our ai-utils package, which you can find on GitHub. With both embedded function calling and our ai-utils, you’re one step closer to creating intelligent AI agents, and from there, the possibilities are endless.

Why Cloudflare’s AI platform?

OpenAI has been the gold Continue reading

NVIDIA Air Infrastructure Simulation Platform

I recently tested the NVIDIA Air Infrastructure Simulation Platform and would like to share my first experiences with you. What is NVIDIA Air Infrastructure Simulation Platform or NVIDIA Air? In a nutshell, NVIDIA Air is a cloud-hosted, data center simulation platform. Where you can: Test and validate network configurations, features, and automation code. Build your own data center topology or choose from an impressive list of pre-built topologies. You can use Cumulus Linux or SONiC as network operating system, add Ubuntu nodes, and more. Import / Export lab topologies. Share the…

The post NVIDIA Air Infrastructure Simulation Platform appeared first on AboutNetworks.net.

AI/ML Networking Part I: RDMA Basics

Remote Direct Memory Access - RDMA Basics


Introduction

Remote Direct Memory Access (RDMA) architecture enables efficient data transfer between Compute Nodes (CN) in a High-Performance Computing (HPC) environment. RDMA over Converged Ethernet version 2 (RoCEv2) utilizes a routed IP Fabric as a transport network for RDMA messages. Due to the nature of RDMA packet flow, the transport network must provide lossless, low-latency packet transmission. The RoCEv2 solution uses UDP in the transport layer, which does not handle packet losses caused by network congestion (buffer overflow on switches or on a receiving Compute Node). To avoid buffer overflow issues, Priority Flow Control (PFC) and Explicit Congestion Notification (ECN) are used as signaling mechanisms to react to buffer threshold violations by requesting a lower packet transfer rate.

Before moving to RDMA processes, let’s take a brief look at our example Compute Nodes. Figure 1-1 illustrates our example Compute Nodes (CN). Both Client and Server CNs are equipped with one Graphical Processing Unit (GPU). The GPU has a Network Interface Card (NIC) with one interface. Additionally, the GPU has Device Memory Units to which it has a direct connection, bypassing the CPU. In real life, a CN may have several GPUs, each with multiple memory units. Intra-GPU communication within the CN happens over high-speed NVLinks. The connection to remote CNs occurs over the NIC, which has at least one high-speed uplink port/interface.

Figure 1-1 also shows the basic idea of a stacked Fine-Grained 3D DRAM (FG-DRAM) solution. In our example, there are four vertically interconnected DRAM dies, each divided into eight Banks. Each Bank contains four memory arrays, each consisting of rows and columns that contain memory units (transistors whose charge indicates whether a bit is set to 1 or 0). FG-DRAM enables cross-DRAM grouping into Ranks, increasing memory capacity and bandwidth.

The upcoming sections introduce the required processes and operations when the Client Compute Node wants to write data from its device memory to the Server Compute Node’s device memory. I will discuss the design models and requirements for lossless IP Fabric in later chapters.



Figure 1-1: Fine-Grained DRAM High-Level Architecture.
Continue reading

DNS Evolution

The DNS is a crucial part of the Internet's architecture. However, the DNS is not a rigid and unchanging technology. It has changed considerably over the lifetime of the Internet and here I’d like to look at what’s changed and what’s remained the same.

Explore and Fix BGP Wedgies

RFC 4264 defines BGP wedgies as “a class of BGP configurations for which there is more than one potential outcome, and where forwarding states other than the intended state are equally stable.” Even worse, “the stable state where BGP converges may be selected by BGP in a non-deterministic manner.”

Want to know more? You can explore a real-life BGP wedgie and fix it in the latest BGP lab exercise.

Explore and fix a BGP wedgie

Automatically replacing polyfill.io links with Cloudflare’s mirror for a safer Internet

polyfill.io, a popular JavaScript library service, can no longer be trusted and should be removed from websites.

Multiple reports, corroborated with data seen by our own client-side security system, Page Shield, have shown that the polyfill service was being used, and could be used again, to inject malicious JavaScript code into users’ browsers. This is a real threat to the Internet at large given the popularity of this library.

We have, over the last 24 hours, released an automatic JavaScript URL rewriting service that will rewrite any link to polyfill.io found in a website proxied by Cloudflare to a link to our mirror under cdnjs. This will avoid breaking site functionality while mitigating the risk of a supply chain attack.

Any website on the free plan has this feature automatically activated now. Websites on any paid plan can turn on this feature with a single click.

You can find this new feature under Security ⇒ Settings on any zone using Cloudflare.

Contrary to what is stated on the polyfill.io website, Cloudflare has never recommended the polyfill.io service or authorized their use of Cloudflare’s name on their website. We have asked them to remove the Continue reading

Cloudflare incident on June 20, 2024

On Thursday, June 20, 2024, two independent events caused an increase in latency and error rates for Internet properties and Cloudflare services that lasted 114 minutes. During the 30-minute peak of the impact, we saw that 1.4 - 2.1% of HTTP requests to our CDN received a generic error page, and observed a 3x increase for the 99th percentile Time To First Byte (TTFB) latency.

These events occurred because:

  1. Automated network monitoring detected performance degradation, re-routing traffic suboptimally and causing backbone congestion between 17:33 and 17:50 UTC
  2. A new Distributed Denial-of-Service (DDoS) mitigation mechanism deployed between 14:14 and 17:06 UTC triggered a latent bug in our rate limiting system that allowed a specific form of HTTP request to cause a process handling it to enter an infinite loop between 17:47 and 19:27 UTC

Impact from these events were observed in many Cloudflare data centers around the world.

With respect to the backbone congestion event, we were already working on expanding backbone capacity in the affected data centers, and improving our network mitigations to use more information about the available capacity on alternative network paths when taking action. In the remainder of this blog post, we will go into Continue reading

Looking for a Simple Multihop EBGP Use Case

I plan to add several challenge labs using multihop EBGP sessions to the BGP labs project, including:

  1. Running BGP between VMs and central BGP route servers
  2. Using multihop EBGP session to send full Internet routing table to a customer without overloading the PE-router
  3. Running EBGP EVPN session between loopbacks advertised with EBGP IPv4 session (🤢)

However, I would love to start with a simple use case to help engineers unfamiliar with BGP realize when they might have to use multihop EBGP sessions. Unfortunately, I can’t find one, and the scenarios where I used multihop EBGP in the past (EBGP load balancing and using a low-end router in the EBGP path, where I was effectively using the reverse application of #2 as a customer) are mostly irrelevant.

Would you have an easy-to-understand use case that is best solved with a multihop EBGP session? Please share it in the comments. Thanks a million!

1 … 135 136 137 138 139 … 3,849