The OSI model has eight layers, right? So why do we spend all our time working on two or three of them? Lee Badman tells us a story about how we need to remember to respect all the layers of the networking model properly.
Consistent delivery is hard to pull off, whether you’re a delivery company or a networking company. With the help of the Ethernet team at Intel, however, you can accomplish some great things and set yourself up for success. Tom Hollingsworth takes a look at Intel’s Application Device Queue technology and how it can help ensure consistent, reliable packet delivery.
Ever wondered why the MTU for Ethernet is 1500 bytes? Ivan Pepelnjak teaches us a bit about why we have this relic of the early days of networking still hanging around in our lives today.
News came out today in The Marker that Microsoft is looking into an acquisition of Mellanox. This would likely mark a major shift to in-house hardware development for Azure, and could form the basis of a composable cloud offering in Microsoft’s cloud. But it would also take a major (if un-heralded) player out of the market for everyone else.
The future of Ethernet is getting faster every day. Andy Bechtolsheim of Arista knows that as well as anyone. That’s why he takes a look at where 400Gbps Ethernet is today and how soon it will be arriving. Tom Hollingsworth looks at his presentation and discusses why it’s a very telling look at the future of networking.
In response to a reader question on his look at Liqid’s composable infrastructure, Russ White frames an interesting question: is it easier to extend PCIe to support switching, and longer runs, or is it easier to design an entire protocol to (effectively) run PCIe over Ethernet? Liqid developed their solution based on former, but other composable infrastructure projects prefer an Ethernet based approach. It’s an interesting look into the benefits and drawbacks of both.
In the enterprise, it’s been interesting to follow the debates between 10/40/100GbE and the alternative 25/50/100GbE roadmaps. As the data center demands more bandwidth, we’ll see this debate shake out in practice. But those kinds of speeds are completely irrelevant to the needs of consumers in any kind of foreseeable future. We’re just really starting to see use of the now ubiquitous 1GbE that’s standard on most devices. That’s what made the announcement that Aquantia is launching a consumer focused line of chips with NBASE-T support intriguing.
Last year I went through my own Mac migration. My wife’s ancient 2006 MacBook in lovely white polycarbonate had a good long life, but was just about becoming unusable. With a maxed out 2GB of RAM and a Core Duo (not a typo) processor, I was actually impressed how long it was relatively functional. This was […]
DriveScale wants to change how storage is considered in your datacenter. Think about how storage is added to a typical setup. If you need more storage on-prem, you throw a couple of pizza boxes on the rack, adding storage, but also compute, memory, connectivity. That’s great, if you just happen to need your storage to scale according to your vendor’s specification.
Surprise! No amount of networking technology will make Layer-2 networks be the correct choice for everything they’re being pitched for.