Hacker News with Generative AI: Infrastructure

Canada pledges billions for high-speed rail linking Quebec City and Toronto (apnews.com)
The Canadian government said Wednesday it is moving ahead with a multibillion-dollar plan to build a high-speed rail network between Quebec City and Toronto.
Putin's secret weapon: The threat to the cables on our sea beds (bbc.co.uk)
Shortly after midday local time on Christmas Day 2024, workers at the Finnish electricity company Fingrid noticed the main undersea electricity cable linking Finland with Estonia was damaged, significantly reducing Estonia’s power supply.
Subsea fibre optic cable deliberately cut for the 2nd time between N.S. and N.L (cbc.ca)
Telecommunications giant Bell is exploring surveillance options in the Gulf of St. Lawrence after one of its subsea fibre optic cables between Cape Breton Island and Newfoundland's west coast was recently severed for the second time.
Trump Administration Moves to End New York's Congestion Pricing Tolls (nytimes.com)
President Trump intends to revoke federal approval of New York City’s congestion pricing program, fulfilling a campaign promise to reverse the policy that tolls drivers who enter Manhattan’s busiest streets to finance repairs to mass transit.
Agent-Less System Monitoring with Elixir Broadway (opsmaru.com)
We've started working on something that will be a critical component the Opsmaru platform. This part requires having proper infrastructure monitoring for several reasons. While Opsmaru has health monitoring for every cluster managed by the platform, it doesn't give us deep insights into the metrics of the systems.
Meta plans to link US and India with longest undersea cable project (theguardian.com)
Meta has announced plans to build the world’s longest underwater cable project, which aims to connect the US, India, South Africa, Brazil and other regions.
Interactive tool shows what projects lawmakers announced in your neighborhood (grist.org)
By the time President Donald Trump retook office, lawmakers had announced nearly $700 billion in funding for infrastructure- and climate-related projects under two bills passed during Joe Biden’s administration — the Inflation Reduction Act and the bipartisan infrastructure law.
Turning Down Upspin Infrastructure (groups.google.com)
Upspin, like Unix and Plan 9, was intended to foster communities of sharing, but has been less successful at that than we hoped. As a consequence, with regret, we have decided to turn down the central infrastructure such as the keyserver over the coming months
Meta’s Hyperscale Infrastructure: Overview and Insights (cacm.acm.org)
Hyperscalers, such as Alibaba, Amazon, ByteDance, Google, Meta, Microsoft, and Tencent, have developed planetary-scale infrastructure to deliver cloud, Web, or mobile services to their global users.
Estonian, Latvian and Lithuanian Power Systems Disconnected from Russian IPS/Ups (elering.ee)
The electricity systems of the Baltic States have been successfully disconnected from the Russia-controlled IPS/UPS system and are now operating independently in island mode.
Meta's Hyperscale Infrastructure: Overview and Insights (cacm.acm.org)
Hyperscalers, such as Alibaba, Amazon, ByteDance, Google, Meta, Microsoft, and Tencent, have developed planetary-scale infrastructure to deliver cloud, Web, or mobile services to their global users.
Grafana: Why observability needs FinOps, and vice versa (grafana.com)
Observability tools have changed the way we monitor infrastructure and applications, as teams get complete visibility into performance across complex, multi-cloud environments.
Trump Puts California High-Speed Rail in Crosshairs (ktla.com)
President Donald Trump has renewed his efforts to stifle California’s High-Speed Rail project that is currently under construction and would link the Bay Area with greater Los Angeles.
Farewell potholes? UK team invents self-healing road surface (theguardian.com)
For all motorists, but perhaps the Ferrari-collecting rocker Rod Stewart in particular, it will be music to the ears: researchers have developed a road surface that heals when it cracks, preventing potholes without a need for human intervention.
Seeking Support After Equinix Metal Sunsets (alpinelinux.org)
We are deeply grateful to Equinix Metal (formerly Packet.net) for their longstanding support, which has been essential to Alpine Linux’s ecosystem. However, with Equinix sunsetting their bare-metal hosting service, this critical support will soon end. Their contributions have been invaluable, and we thank them for helping keep Alpine Linux reliable and efficient.
Alpine Linux: Seeking Support After Equinix Metal Sunsets (alpinelinux.org)
We are deeply grateful to Equinix Metal (formerly Packet.net) for their longstanding support, which has been essential to Alpine Linux’s ecosystem. However, with Equinix sunsetting their bare-metal hosting service, this critical support will soon end. Their contributions have been invaluable, and we thank them for helping keep Alpine Linux reliable and efficient.
Norway seizes Russian-crewed ship on suspicion of damage to undersea cable (cnn.com)
Github Incident with Pull Requests and Issues (githubstatus.com)
Users may experience timeouts in various GitHub services. We have identified an issue with our caching infrastructure and are working to mitigate the issue
Real-time, river sections downstream of sewage discharges from storm overflows (sewagemap.co.uk)
Another undersea cable damaged in Baltic Sea (france24.com)
Another undersea data cable, this time connecting Sweden and Latvia, has been severed in the Baltic Sea, officials from both countries said Sunday, and prompted Latvia to dispatch a warship near to where the incident had occurred.
How we scaled Slack to support 1000s of developers (railway.com)
Railway makes software infrastructure for humans. Our pitch is simple. You give us a docker image or GitHub repo. We deploy and scale it, no friction.
The first release of OpenVox, the OSS implementation of Puppet (overlookinfratech.com)
It’s been quite a journey, y’all. But we’re excited to announce the first release of OpenVox, the community-maintained open source implementation of Puppet. OpenVox 8.11 is functionally equivalent to Puppet and should be a drop-in replacement. Be aware, of course, that even though you can type the same commands, use all the same modules and extensions, and configure the same settings, OpenVox is not yet tested to the same standard that Puppet is.
Trump announces a $500B private AI infrastructure investment (cnn.com)
Trump to announce private sector AI infrastructure investment (cbsnews.com)
President Trump is set to announce billions of dollars in private sector investment to build artificial intelligence infrastructure in the United States, CBS News has learned.
Show HN: Pica – Rust-based agentic AI infrastructure (open-source) (picaos.com)
Bill Banning Foreign Adversaries Setting Up Shop Near Critical Areas Advances (cowboystatedaily.com)
The Senate Appropriations Committee gave a unanimous thumbs up Thursday to a bill that would ban foreign adversaries from setting up near critical infrastructure in Wyoming.
Driving blind: NYC subways steered by 1930s tech, paper maps and a lot of hope (gothamist.com)
Penn Station Can Handle the Load: New York Is Ready for Through-Running (etany.org)
The recently funded Gateway Program, which centers around building two new rail tunnels under the Hudson to connect New Jersey to Manhattan's Pennsylvania Station, will enable a massive increase in rail passenger traffic. To handle this growth, ETA has continually advocated for through-running trains from New Jersey to Queens and beyond.
Enterprises in for a shock when they realize power and cooling demands of AI (theregister.com)
Most businesses rushing to adopt AI are unprepared for the energy demands it'll place on their infrastructure, and few have a handle on the power consumption of AI systems or the implications for their datacenters.
Scaling LLMs with Golang: How we serve millions of LLM requests (assembled.com)
While the LLM ecosystem is overwhelmingly Python-first, we've found Go to be exceptionally well-suited for production deployments. Our Go-based infrastructure handles millions of monthly LLM requests with minimal performance tuning. Beyond Go's well-documented advantages (see Rob Pike’s excellent distillation of Go's benefits), three capabilities have proven particularly valuable for LLM workloads: static type checking for handling model outputs, goroutines for managing concurrent API calls, and interfaces for building composable response pipelines.