Category: switching

Years ago I managed to saturate a 10GE uplink on a vSphere server I tested with a single Linux VM using less than one vCPU. On the other hand, squeezing 1 Gbps out of Open vSwitch using GRE encapsulation was called ludicrous speed not so long ago. Implementing overlay virtual networking in the hypervisor obviously carries a huge performance penalty, right? Not so fast…

In the Myths That Refuse to Die: Scalability of Overlay Virtual Networking blog post I wrote “number of MAC addresses has absolutely no impact on the forwarding performance until the MAC hash table overflows”, which happens to be almost true.

Industry press, networking blogs, vendor marketing whitepapers and analyst reports are full of grandiose claims of benefits of whitebox switching and hardware disaggregation. Do you ever wonder whether these people actually practice their theories?

One of the networking engineers using my ExpertExpress to validate their network design had an interesting problem: he was building a multi-tenant VLAN-based private cloud architecture with each tenant having multiple subnets, and wanted to route within the tenant network as close to the VMs as possible (in the ToR switch).

He was using Nexus 5600 as the ToR switch, and although there’s conflicting information on the number of VRFs supported by that switch (verified topology: 25 VRFs, verified maximum: 1000 VRFs, configuration guide: 64 VRFs), he thought 25 VRFs (tenant routing domains) might be enough.

Simonp made a perfectly valid point in a comment to my latest OVS blog post:

Obviously the page you're referring to is a quick-and-dirty benchmark. If you wanted the optimal numbers, you would have to tune quite a few parameters just like for hardware benchmarks (sysctl kernel parameters, Jumbo frames, ...).

While he’s absolutely right, this is not the performance data a typical user should be looking for.

One of my readers is struggling with the aftermath of marketing gimmicks:

We will implement a new network soon, and we're discussing P-routers versus regular routers versus switches. I'm looking for arguments to go one way or the other.

TL&DR: there’s no difference between router and L3 switch.

Want to know what the difference between Virtual Chassis and Virtual Chassis Fabric is? How Local Link Bias works? How ISSU on QFX 5100 works even though the box doesn’t have two supervisor boards? You’ll find answers to all these questions in new videos describing Juniper data center switches.

The pilot episode of Software Gone Wild podcast featuring Snabb Switch created plenty of additional queries (and thousands of downloads) – it was obviously time for another deep dive episode discussing the intricate innards of this interesting virtual switch.

During the deep dive Luke Gorrie, the mastermind behind the Snabb Switch, answered a long list of questions, including:

A comment by Pieter E. Smit on my vSphere Does Not Need LAG Bandaids post opened yet another can of worms: vSphere behavior on uplink recovery.

Short summary: vSphere starts using an uplink as soon as its physical layer becomes operational, which might happen during ToR switch startup phase, or before a ToR switch port enters forwarding state.

When describing Hyper-V Network Virtualization packet forwarding I briefly mentioned that the hypervisor switches create (an equivalent of) a host route for every VM they need to know about, prompting some readers to question the scalability of such an approach. As it turns out, layer-3 switches did the same thing under the hood for years.

The SDN industry probably considers me an old and grumpy naysayer (and I’m positive Mrs Y has a special place in their hearts after her recent blog post), so I tried really hard to find a real-life example where OpenFlow could be used to solve mid-market innovator’s dilemma to balance my usual OpenFlow and SDN presentation.

It took years before the rumored Cisco vSwitch materialized (in the form of Nexus 1000v), several more years before there was the first competitor (IBM Distributed Virtual Switch), and who knows how long before the third entrant (recently announced HP vSwitch) jumps out of PowerPoint slides and whitepapers into the real world.

Compare that to the Hyper-V environment, where we have at least two virtual switches (Nexus 1000V and NEC's PF1000) mere months after Hyper-V's general availability.

I’m probably flogging a fossilized skeleton of a long-dead horse, but it seems I never wrote about this topic before, so here it is (and you might want to read this book for more details).

Process switching is the oldest, simplest and slowest packet forwarding mechanism. Packets received on an interface trigger an interrupt, the interrupt handler identifies the layer-3 protocol based on layer-2 packet headers (example: Ethertype in Ethernet packets) and queues the packets to (user mode) packet forwarding processes (IP Input and IPv6 Input processes in Cisco IOS).

Autumn must be a perfect time for data center product launches: last week Brocade launched its core VDX switch and yesterday Arista and Cisco launched their new low-latency switches (yeah, the simultaneous launch must have been pure coincidence).

I had the opportunity to listen to Cisco’s and Arista’s product briefings, continuously experiencing a weird feeling of déjà vu. The two switches look like twin brothers … but there are some significant differences between the two:

Arista is launching a new product line today shrouded in mists of SDN and cloud buzzwords: the 7150 series top-of-rack switches. As expected, the switches offer up to 64 10GE ports with wire speed L2 and L3 forwarding and 400 nanosecond(!) latency.

Also expected from Arista: unexpected creativity. Instead of providing a 40GE port on the switch that can be split into four 10GE ports with a breakout cable (like everyone else is doing), these switches group four physical 10GE SFP+ ports into a native 40GE (not 4x10GE LAG) interface.

But wait, there’s more...