ChatGPT’s rivals, Kwai’s quiet rise: the top Internet services of 2025

In 2025, the Internet is more central to our lives than ever, and we rely on an array of online services to get things done, connect with others, and enjoy ourselves. Cloudflare’s Top Internet Services of 2025 report explores how the connected world interacted this year, based on Cloudflare’s observations and analysis of DNS trends. 

This report is part of the 2025 Cloudflare Radar Year in Review, focused on shifts in popularity of Internet services. We hope you find the results are a compelling view of trends in nine major categories — who’s moving up, who’s sliding down, and who continues to hold our attention.

These rankings show relative popularity within each category, based on anonymized DNS query data from Cloudflare’s 1.1.1.1 DNS resolver and a machine-learning-assisted ranking method introduced in 2022. A lower rank does not imply lower traffic, only that other services may have grown faster.

Categories

  • Generative AI ➜

  • Social Media ➜

  • E-commerce ➜

  • Video Streaming ➜

  • News ➜

  • Messaging ➜

  • Metaverse & Gaming ➜

  • Financial Services ➜

  • Cryptocurrency Services ➜

Key trends and takeaways

From the dominance of social media and streaming to the rapid growth of AI chatbots, the data Continue reading

Underscores (in Hostnames) Strike Again

I don’t know why I decided to allow underscores in netlab node names. Maybe it’s a leftover from the ancient days when some network devices refused to accept hyphens in hostnames, or perhaps it’s a programmer’s subconscious hatred of hyphens in identifiers (no programming language I’m aware of allows them for a very good reason).

Regardless, you can use underscores in netlab node names (and plugins like multilab use them to create unique hostnames), and they work great on Linux distributions we recommend… until they don’t.

What follows is a story about the weird dependencies that might bite you if you ignore ancient RFCs.

Read more …

Otak Anda, Atlet Tersembunyi: Mengungkap Rahasia Neuro-Olahraga yang Jarang Dibahas

Ketika kita membicarakan olahraga, fokus kita seringkali tertuju pada otot yang kuat, teknik yang sempurna, atau strategi brilian di lapangan. Kita memuja atlet karena fisik mereka yang menakjubkan. Tapi, pernahkah terbersit di benak Anda, apa sebenarnya yang membedakan seorang atlet elit dengan mereka yang hanya “baik”? Jawabannya mungkin bukan terletak di otot mereka, melainkan di dalam organ yang paling kompleks di tubuh: otak.

Selamat datang di dunia Neuro-Olahraga, sebuah bidang yang masih tergolong baru namun revolusioner. Ini adalah ilmu yang mempelajari koneksi langsung antara sistem saraf, fungsi kognitif, dan performa fisik. Artikel ini akan mengajak Anda melampaui pembahasan “fokus” atau “konsentrasi” yang klise, dan menyelami mekanisme otak yang sesungguhnya bekerja di balik setiap gerakan presisi, setiap keputusan sepersekian detik, dan setiap momen “keajaiban” dalam olahraga.

Apa Itu Neuro-Olahraga? Lebih dari Sekadar Mental Blocking

Neuro-olahraga bukanlah psikologi Continue reading

Lab: Multilevel IS-IS Deployments

Like OSPF, IS-IS was designed when router memory was measured in megabytes and clock speeds in megahertz. Not surprisingly, it includes a scalability mechanism similar to OSPF areas. An IS-IS router could be a level-1 router (having in-area prefixes and a default route), a level-2 router (knowing just inter-area prefixes), or a level-1-2 router (equivalent to OSPF ABR).

Even though multilevel IS-IS is rarely used today, it always makes sense to understand how things work, and the Multilevel IS-IS Deployments lab exercise created by Dan Partelly gives you a perfect starting point.

Click here to start the lab in your browser using GitHub Codespaces (or set up your own lab infrastructure). After starting the lab environment, change the directory to advanced/1-multilevel and execute netlab up.

An In-Depth Look at Istio Ambient Mode with Calico

The Next Step Toward a Unified Kubernetes Platform: Istio Ambient Mode

Organizations are struggling with rising operational complexity, fragmented tools, and inconsistent security enforcement as Kubernetes becomes the foundation for modern application platforms. As a result of this complexity and fragmentation, platform teams are increasingly burdened by the need to stitch together separate solutions for networking, network security, and observability. This fragmentation also creates higher operating costs, security gaps, inefficient troubleshooting, and an elevated risk of outages in mission-critical environments. The challenge is even greater for companies running multiple Kubernetes distributions, as relying on each platform’s unique and often incompatible networking stack can lead to significant vendor lock-in and operational overhead.

The Tigera Unified Strategy: Addressing Fragmentation

Tigera’s unified platform strategy is designed to address these challenges by providing a single solution that brings together all the essential Kubernetes networking and security capabilities enterprises need, that includes Istio Ambient Mode, delivered consistently across every Kubernetes distribution.

Istio Ambient Mode brings sidecarless service-mesh functionality that includes authentication, authorization, encryption, L4/L7 traffic controls, and deep application-level (L7) observability directly into the unified Calico platform. By including Istio Ambient Mode with Calico and making it easy to install and manage with the Tigera Continue reading

React2Shell and related RSC vulnerabilities threat brief: early exploitation activity and threat actor techniques

On December 3, 2025, immediately following the public disclosure of the critical, maximum-severity React2Shell vulnerability (CVE-2025-55182), the Cloudforce One Threat Intelligence team began monitoring for early signs of exploitation. Within hours, we observed scanning and active exploitation attempts, including traffic originating from infrastructure associated with Asian-nexus threat groups.

Early activity indicates that threat actors quickly integrated this vulnerability into their scanning and reconnaissance routines. We observed systematic probing of exposed systems, testing for the flaw at scale, and incorporating it into broader sweeps of Internet‑facing assets. The identified behavior reveals the actors relied on a combination of tools, such as standard vulnerability scanners and publicly accessible Internet asset discovery platforms, to find potentially vulnerable React Server Components (RSC) deployments exposed to the Internet.

Patterns in observed threat activity also suggest that the actors focused on identifying specific application metadata — such as icon hashes, SSL certificate details, or geographic region identifiers — to refine their candidate target lists before attempting exploitation. 

In addition to React2Shell, two additional vulnerabilities affecting specific RSC implementations were disclosed: CVE-2025-55183 and CVE-2025-55184. Both vulnerabilities, while distinct from React2Shell, also relate to RSC payload handling and Server Function semantics, and are described in more detail Continue reading

IETF v6ops Working Group with Nick Buraglio

The first IPv6 specs were published in 1995, and yet 30 years later, we still have a pretty active IETF working group focused on “developing guidelines for the deployment and operation of new and existing IPv6 networks.” (taken from the old charter; they updated it in late October 2025). Why is it taking so long, and what problems are they trying to solve?

Nick Buraglio, one of the working group chairs, provided some answers in Episode 203 of the Software Gone Wild podcast.

Listen to the podcast

Driving HPC Performance Up Is Easier Than Keeping The Spending Constant

We are still mulling over all of the new HPC-AI supercomputer systems that were announced in recent months before and during the SC25 supercomputing conference in St Louis, particularly how the slew of new machines announced by the HPC national labs will be advancing not just the state of the art, but also pushing down the cost of the FP64 floating point operations that still drives a lot of HPC simulation and modeling work. …

Driving HPC Performance Up Is Easier Than Keeping The Spending Constant was written by Timothy Prickett Morgan at The Next Platform.

TCG065: AutoCon 4 Recap, AI Tools, MCP’s First Birthday, and More

In this year-end episode, William and Eyvonne recap their experiences at AutoCon 4 in Austin, Texas. They discuss the conference’s new multi-track format, including Eyvonne’s presentation in the leadership track on why technical projects fail. The conversation dives into how AI tools like Google Gemini can augment – not replace – human creativity, from research... Read more »

UET Protocol: How the NIC constructs packet from the Work Entries (WRE+SES+PDS)

 Semantic Sublayer (SES) Operation 

[Rewritte 12. Dec-2025]

After a Work Request Entity (WRE) is created, the UET provider generates the parameters needed by the Semantic Sublayer (SES) headers. At this stage, the SES does not construct the actual wire header. Instead, it provides the header parameters, which are later used by the Packet Delivery Context (PDC) state machine to construct the final SES wire header, as explained in the upcoming PDC section. These parameters ensure that all necessary information about the message, including addressing and size, is available for later stages of processing.

Fragmentation Due to Guaranteed Buffer Limits

In our example, the data to be written to the remote GPU is 16 384 bytes. The dual-port NIC in figure 5-5 has a total memory capacity of 16 384 bytes, divided into three regions: a 4 096-byte guaranteed per-port buffer for Eth0 and Eth1, and an 8 192-byte shared memory pool available to both ports. Because gradient synchronization requires lossless delivery, all data must fit within the guaranteed buffer region. The shared memory pool cannot be used, as its buffer space is not guaranteed.

Since the message exceeds the size of the guaranteed buffer, it must be fragmented. The UET provider splits the 16 384-byte Continue reading

D2DO289: Instana: Leading the Future of Observability (Sponsored)

As AI tools and agentic AI become part of how applications are developed, delivered, and managed, application performance monitoring and observability have to adapt. Ned Bellavance sits down with Drew Flowers and Jacob Yackenovich from IBM Instana about where these fields sit today, and the potential impacts of AI. They detail the challenges of application... Read more »

Why AI Traffic Growth Demands Optical Network Automation Now

Demand for bandwidth continues to surge with increasing use of AI, high-definition video streaming, cloud and 5G mobile applications. In response, optical services need to become more dynamic and meet stricter service-level requirements. Consequently, metro-based networks owned by communications service providers (CSPs) and used for mobile transport, enterprise services or broadband access services face new demands relating to power, multiservice aggregation, security, AI integration and network slicing. Figure 1 shows a rich set of applications in a typical metro network. Figure 1 — A typical metro network supports a rich set of applications. As AI use grows, it will create a force multiplier that will drive significant traffic to transport networks. But the challenging throughput, latency and reliability requirements of AI workloads are already taking network scale and complexity to another level. CSPs are facing unprecedented pressure to deliver more, faster and better. Figure 2 shows the impact of AI traffic growth based on a recent Nokia Bell Labs study. This growth can range from 14% to 31%. Figure 2 — Traffic growth effects on CSP networks. CSPs’ access and aggregation network (orange) will bear the heaviest load, roughly 31% of the total AI traffic. CSPs’ metro network (pink) is Continue reading

Simplifying SR-TE with mesh templates

A big problem with deploying traffic engineering is configuration complexity. This has now been solved.

Why use Traffic Engineering

Typical reasons for using traffic engineering are:

  • Send some traffic via a low latency path
  • Send A/B streams via different paths
…
1 … 3 4 5 6 7 … 3,836