AI projects don’t fail because models don’t work or GPUs lack performance. …
Attending GTC? Join Us For An Exclusive Roundtable Dinner On AI Data Platforms was written by Atul Chaudhary at The Next Platform.
netlab release 26.02 is out, including the usual potpourri of goodies:
The fun part, however, are the new container configuration methods:
In the modern AI datacenter – really, a data galaxy at this point because AI processing needs have broken well beyond the bounds of a single datacenter or even multiple datacenters in a region in a few extreme cases – has two pinch points in the network. …
Cisco Doubles Up The Switch Bandwidth To Take On AI Scale Up And Scale Out was written by Timothy Prickett Morgan at The Next Platform.
The NVIDIA GTC conference has a reputation for delivering announcements that reshape industry roadmaps. …
The Greatest AI Show On Earth was written by David Gordon at The Next Platform.

In the previous posts in this series, we covered the basics of multicast, IGMP, PIM Dense Mode, and PIM Sparse Mode. In the Sparse Mode post, we manually configured the RP address on every router in the network. This works fine in a small lab, but in a larger network with many routers, it becomes difficult to manage. If the RP changes, you have to update the configuration on every single router.

AutoRP solves this problem by allowing routers to dynamically learn the RP address. Instead of manually configuring the RP on each router, you configure one or more routers to announce themselves as Candidate RPs. A separate router (or the same as the Candidate RP router) called the Mapping Agent collects these announcements and distributes the RP information to all other routers in the network. This makes RP management much easier and also provides a way to implement RP redundancy.
There are two methods to dynamically learn the RP address, which are Auto-RP and Bootstrap Router (BSR). In this Continue reading
Brian Linkletter published an updated overview of open-source network simulators and emulators.
containerlab and GNS3 are clear leaders (no surprise there) with the original vrnetlab becoming abandonware (fortunately, we have Roman Dodin’s fork), which makes me think we should focus on using netlab primarily with containerlab and slowly sunset the Vagrant support, particularly considering some people actively hate the license change.
Also, if anyone feels like writing an interface (provider module) between netlab and GNS3, the pull request would be most welcome 😎
Any thoughts? Please leave a comment!
This is turning into a “dog bites man” story, but the forecasts for spending in the datacenter for this year keep going up and up, and a few days ago Gartner’s economists and prognosticators finished up their tea and looked at the leaves at the common of a cup through a polished crystal ball and predicted that datacenter spending this year would go up. …
Datacenter Spending Forecast Revised Upwards – Yet Again was written by Timothy Prickett Morgan at The Next Platform.
After the enormous speedup I achieved with the FRR containers, I tried to do something similar with the Arista cEOS ones. After all, Arista’s pretty open about running its software on standard Linux, so it should be possible to map host-side configuration files into container-side scripts and execute them, right?
There was just one tiny gotcha: all netlab-generated EOS configuration files are device configuration snippets that are intended to be submitted via EOS CLI, and I didn’t feel like cracking open the netmiko documentation (that’s another backburner project).
However, Arista cEOS includes this magic command called FastCli ;)
Like Google and Meta Platforms, Amazon knows exactly how to infuse AI into its business operations such as online retail, transportation, advertising, and even the Amazon Web Services cloud. …
The Twin Engine Strategy That Propels AWS Is Working Well was written by Timothy Prickett Morgan at The Next Platform.