Welcome!

Rethinking the Network

Marten Terpstra

Subscribe to Marten Terpstra: eMailAlertsEmail Alerts
Get Marten Terpstra via: homepageHomepage mobileMobile rssRSS facebookFacebook twitterTwitter linkedinLinkedIn


Top Stories by Marten Terpstra

Much has been published about the Open Compute Project. Initiated by Facebook, it has become an industry effort focused on standardization of many parts and components in the datacenter. Initially focused on racks, power and server design, it has also added storage and now networking to its fold. Its goal is fairly straightforward: “how can we design the most efficient compute infrastructure possible”, a direct quote from its web site. The focus of OCP has been mostly around hardware designs and specifications. If you look at the networking arm of OCP, you find several Top of Rack (ToR) ethernet switch hardware designs donated by the likes of Broadcom, Mellanox and Intel. By creating open specifications of hardware designs for fairly standard ethernet switches, the industry can standardize on these designs and economics of scale would drive down the cost to create an... (more)

Changing the Way We Configure and Provision Our Networks

Some people believe good or bad things always happen in threes. I believe you will always be able to find three (and probably more) things that are good or bad and somewhat related, but sometimes I get surprised by the apparent coincidental appearance of several closely related “things”. Last week the folks at networkheresy.com posted a second installment of their “policy in the datacenter” discussion, Cisco announced the acquisition of tail-f and internal to Plexxi we had several intense architectural discussions around Configuration, Provisioning and Policy management. Maybe we... (more)

Scripting Is Automation, But Automation Is Not Scripting

Last week Greg Ferro (@etherealmind) wrote this article about his experience with scripting as a method for network automation, with the ultimate conclusion that scripting does not scale. Early in my career I managed a small network that grew to be a IP over X.25 hub of Europe for a few years providing many countries with their first Internet connectivity. Scripts were everywhere, small ones to grab stats and create pretty graphs, others that continuously checked the status of links and would send emails when things went wrong. While it is hard to argue with Greg’s complaints per... (more)

Complexity Kills Networks

In reviewing some opportunities for Plexxi this week, I was reminded that we have made things very hard on ourselves. Through no one’s fault but our own we have created monsters of networks that are impossible to maintain, debug, diagnose and understand. I have been lucky in my career. Most of it has been in an R&D organization for a large network vendor, but always in positions where I was close to customers, close to customer networks, and of course with that comes the “close to customer networks that are not working well”. And while there are always exceptions, in most of the... (more)

Attention: Overlay Tunnel Construction Ahead

A while ago I wrote a few articles describing the various tunnel protocols used for network virtualization between vSwitches on servers, and between vSwitches and physical network gateways. These are the mechanisms that construct overlay networks on top of a physical network. VMWare uses STT as the tunneling mechanism between vSwitches on servers and VXLAN to communicate with gateways to the non virtualized world. NVGRE is used mostly by Microsoft, and is an extension to GRE tunneling that has been around for a while. Each one of these mechanisms have their pros and cons. They ... (more)