Category: Switching

We did several podcasts describing how one could get stellar packet forwarding performance on x86 servers reimplementing the whole forwarding stack outside of kernel (Snabb Switch) or bypassing the Linux kernel and moving the packet processing into userspace (PF_Ring).

Now let’s see if it’s possible to improve the Linux kernel forwarding performance. Thomas Graf, one of the authors of Cilium claims it can be done and explained the intricate details in Episode 64 of Software Gone Wild.

One of the common taglines parroted by SDN aficionados goes along the lines of “The cost to acquire and manage server and storage architectures has declined over time while networking stays stubbornly expensive.” (I took it straight from an anonymous blog comment).

Let’s see how well it matches reality.

When LinkedIn announced their Project Falco I knew exactly what one of my future Software Gone Wild podcasts would be: a chat with Russ White (Mr. CCDE, now network architect @ LinkedIn).

It took us a long while (and then the summer break intervened) but I finally got it published: Episode 62 is waiting for you.

One of my readers has customers that already experienced performance challenges with Tomahawk-based data center switches. He sent me an email along these lines:

My customers are concerned about buffer performance for packets that are 200 bytes and under. MORE IMPORTANTLY, a customer informed me that there were performance issues when running 4x25GE connections when one group of ports speaks to another group.

Reading the report Mellanox published not so long ago it seems there really is something fishy going on with Tomahawk.

Russ White made an excellent remark while discussing the news that the CloudRouter pushed 650 Gbps through commodity hardware: “If this is software defined networking, then we’ve been doing this since sometime in the 1990’s, perhaps even earlier…”

He’s absolutely right – the first routers (like AGS or IGS from Cisco) did all packet forwarding in software, so as I explained during the Introduction to SDN webinar while reaching dozens of gigabits with software-based packet forwarding is exciting, calling it SDN doesn’t make much sense.

Some people conflate SDN with whitebox switches preferably running Linux. So what exactly is software-hardware disaggregation, and how do whitebox switches and third-party network operating systems fit into the bigger picture?

I tried to answer these questions in the SDN is not whitebox switching part of (free) Introduction to SDN webinar.

Imagine you want to have an IPv6-only access network and transport residual IPv4 traffic tunneled across it. Sounds great, but you need to terminate those tunnels and encapsulate/decapsulate IPv4 traffic at multi-gigabit rate.

There are plenty of reassuringly-expensive hardware solutions that can do that, or you could work with really smart people and get software-based solution that can do 20 Gbps per CPU core.

This article was initially sent to my SDN mailing list. To register for SDN tips, updates, and special offers, click here.

Usman asked a few questions in his comment on my blog, including:

At the moment, local RIB gets downloaded to FIB and we get packet forwarding on a router. If we start evaluating too many fields (PBR) and (assume) are able to push these policies to the FIB - what would become of the FIB table size?

Short answer: It would explode ;)

A lot of people love to talk about ASICs and merchant silicon, but very few really understand the basics. Now there’s a quick way to fix that: watch the excellent Tech Field Day video with Dave Zacks from Cisco Systems.

A few days ago Dell announced their next-generation network OS based on Debian Linux, and bloggers (like my good friend Tom Hollingsworth) started wondering what’s going to happen with Cumulus Linux.

Let’s get into prognostication mode…