The 8th PLNOG meeting starts in less than a week and the fantastic Andrzej Targosz has yet again kindly invited me to talk about cloud networking and data center fabrics (the first session is on Monday afternoon, the second one on Tuesday morning). I’ll be in Warsaw from Sunday evening to late Wednesday morning, so if you’d like to meet me, ask me a really tough question (layer-2 DCI is off-limits), discuss your network design, or just drink a cup of coffee or a beer with me (please don’t mention vodka), send me an e-mail and we’ll figure out where and when to meet.
A reader of my blog sent me this question:
I am a Telecommunication Engineer currently preparing for the CCIE exam. Do you think that in a near future it will be worth to be a CCIE, due to the recent developments like Nicira? What will be the future of Cisco IOS, and protocols like OSPF or BGP? I am totally disoriented about my career.
Well, although I wholeheartedly agree with recent post from Derick Winkworth, the sky is not falling (yet):
I thought Edge Virtual Bridging (EVB) would be the technology transforming the kludgy vendor-specific VM-aware networking solutions into a properly designed architecture, but the launch of L2-over-IP solutions for VMware and Xen hypervisors is making EVB obsolete before it ever made it through the IEEE doors.
After all the laws-of-physics-are-changing hype it must have been anticlimactic for a lot of people to realize what Nicira is doing (although I’ve been telling you that for months). Not surprisingly, there were the usual complaints and twitterbursts:
I finally found just the right set of tools to draw and update webinar roadmaps without too much hassle, and updated all of them to include the webinars developed during late 2011 and planned for 2012:
Tags: Data Center
I didn’t expect we’d see multi-vendor OpenFlow deployment any time soon. NEC and IBM decided to change that and Tervela, a company specialized in building messaging-based data fabrics, decided to verify their interoperability claims. Janice Roberts who works with NEC Corporation of America helped me get in touch with them and I was pleasantly surprised by their optimistic view of OpenFlow deployment in typical enterprise networks.
February 2007 was the month of the static routes: from Reliable static routing to the stupidest idea I ever had: Unequal load split with static routes. I was also checking the administrative distance of various static routes in the Not all static routes were created equal post.
February 2009 started with a Microtik CLI bug that crashed a lot of Cisco routers. I wrote about Oversized AS paths: Cisco IOS bug details, Root cause analysis: oversized AS paths and got into Rant: Priceless mood. We also discussed How useless is UMTS?
We all know that the performance of virtual networking appliances (firewalls, load balancers, routers ... running inside virtual machines) really sucks, right? Some vendors managed to offload the packet-intensive processing into the hypervisor kernel, getting way more bang for the buck, but that’s a pretty R&D-intensive undertaking.
We also know that The Real Men use The Real Hardware (ASICs and FPGAs) to get The Real Performance, right? Wrong!
Let's start with Startup of the Week - Dmitri Kalintsev decided to blog regularly. Make sure you add him to your RSS reader.
And here are the interesting articles accumulated in my Evernote notebook in somewhat random order:
Three days ago IBM launched Distributed Virtual Switch 5000V, its own distributed vSwitch for VMware ESX platform. On one hand, it proves Cisco has been going the right way with Nexus 1000V (just in case you wondered), on the other hand, things just got way more interesting – IBM is obviously returning to networking.
Just before 2011 hit its expiration date, Derick Winkworth published Being Good at IT Stuff where among a gazillion things I totally agree with he also wrote “Even in IT, an IT degree is useless.”
I know exactly why he wrote that; I’d attended plenty of seemingly useless lectures (although it turns out sometimes it pays to understand those topics), and some people still think teaching History of Computer Engineering or obscure programming languages makes perfect sense.
I figured out I wrote a lot about Microsoft Network Load Balancing (NLB) without ever explaining how that marvel of engineering works. To fix that omission, here’s a short video taken from the Data Center 3.0 webinar.
Whenever the networking industry invents a new (somewhat radical) technology, bandwidth-on-demand seems to be one of the much-touted use cases. OpenFlow/SDN is no different – Juniper used its OpenFlow implementation (Open vSwitch sitting on top of Junos SDK) to demonstrate Bandwidth Calendaring (see Dave Ward’s presentation @ OpenFlow Symposium for more details), Greg Ferro was talking about the same topic in his fantastic Introduction to OpenFlow/SDN webinar, and Dmitri Kalintsev recently blogged “How about an ability for things like Open vSwitch ... to actually signal the transport network its connectivity requirements ... say desired bandwidth” I have only one problem with these ideas: I’ve seen them before.
One of my readers sent me the following question:
I know that if I purchase a single webinar, I can apply that cost toward a yearly subscription. Is it additive? If I start buying recordings one-by-one until I reach $200, will I be given the yearly subscription?
The answer is a qualified yes (and you can find all the details on my web site):
Numerous articles published in the last few days describing how Nicira clashes heads-on with Cisco and Juniper just proved that you should never let facts interfere with a good story (let alone eye-catching headline). Just in case you got swayed away by those catchy stories, here’s the real McCoy (as I see it):
I got intrigued when reading Nicira’s white paper claiming their Open vSwitch can run within vSphere/ESX hypervisor. There are three APIs that you could use to get that job done: dvFilter API (intercepting VM NIC like vCDNI does), the API used by Cisco’s Nexus 1000V, or the device driver interface (intercepting uplink traffic). Turns out Nicira decided to use a fourth approach using nothing but publicly-available APIs.
We have a public holiday today, so I’ll spend the morning with my kids instead of writing yet another whatever-does-not-scale post. However, I did stumble across two fantastic cartoons that I simply have to share with you.
My next live webinar will be based on the DMVPN design posts I wrote recently and a number of scenarios that landed in my Inbox during the last few months. I’ll try to help you decide which phase of the DMVPN technology to use, which routing protocol would be best for you, and how to optimize the routing protocol you select. We’ll also discuss interesting redundancy and primary/backup scenarios, including combinations of DMVPN, MPLS/VPN and 3G networks.
Nicira, the OpenFlow startup behind the Open vSwitch, has finally dropped the stealthy cloak. Congratulations!!! Their web site is still pretty sparse on details, but you can get an initial impression of what they’re doing from a number of white papers describing Network Virtualization Platform and DVNI architecture. Short summary: I was almost right, but being a routing-and-switching bloke missed a few interesting bits – OpenFlow (and Open vSwitch) can easily combine security and forwarding functionality.
Two days ago I described how you can use tunneling or labeling to reduce the forwarding state in the network core (which you have to do if you want to have reasonably fast convergence with currently-available OpenFlow-enabled switches). Now let’s see what you can do in the very limited world of OpenFlow 1.0 (if any shipping physical switch supports OpenFlow 1.1 beyond OpenFlow 1.0 functionality, please write a comment)
For whatever reason, Easy Virtual Network (EVN), a configuration sugar-glaze on top of VRF-lite (oops, multi-VRF) that has been lurking in the shadows for the last 18 months erupted into the twittersphere after Cisco’s latest switching launch. I can’t possibly understand why the implementation of a decade-old technology on mature platform (Catalyst 4500 and Catalyst 6500) makes news at the time when 40GE and 100GE interfaces were launched, but the intricacies of marketing always somehow escaped me.
Yesterday I described how the limited flow setup rates offered by most commercially-available switches force the developers of production-grade OpenFlow controllers to drop the microflow ideas and focus on state abstraction (people living in a dreamland usually go in a totally opposite direction). Before going into OpenFlow-specific details, let’s review the existing forwarding state abstraction technologies.