Neat Cisco Nexus Options You May Have Missed

Since Cisco produces plenty of nifty new options, the title would possibly counsel that is going to be a LONG weblog. Removed from it! As I write, it’s Summer time, and I think most readers have higher issues to do than learn lengthy blogs. Brevity for the win!

My purpose right here is to briefly cowl one nifty function that most individuals I’ve talked to will not be conscious of, and one other that I’d by some means missed listening to about. I’ll allow you to guess which one of many following is which.

(Picture: Gorodenkoff/Shutterstock)


Clever Site visitors Director (ITD) offers multi-terabit L4 load balancing and visitors steering by way of Nexus switches. Clustering and symmetric flows are supported. VPC can also be supported. ITD reportedly has a lot much less TCAM use than WCCP.

ITD isn’t a L7 SLB with URL rewrite guidelines, cookie visibility, SSL offload, and many others. Nevertheless, I’ve seen most websites utilizing their NetScaler, F5, or A10 SLB in solely primary VIP methods, so that will suit your wants, or not. As well as, if you happen to want terabit SLB conduct, streamlining how your webpages are accessed to leverage present Nexus switches may be extraordinarily engaging in comparison with a pile of prime finish F5’s or different SLB’s. Not taking a place right here, simply making an attempt to elucidate how and the place ITD would possibly or won’t be helpful.

ITD is supported on the Nexus 5000, 6000, 7000 sequence, and 9000 (broadly talking), with the correct code and license stage (see the At A Look hyperlink under for particulars and double-check pre-requisites and gadget assist for the particular Nexus mannequin you bear in mind).

Apparently ITD auto-configures PBR to implement symmetry of flows (the place related), presumably with failure triggering re-allocation (hopefully simply failed flows, however I’m not seeing that element within the documentation).

Abstract of Modes of Use:

  • Direct Server Return (DSR) mode: the SLB (ITD) forwards to server (normally at L2 by MAC, ITD is configured with IP of interface), VIP is configured on loopback on servers in pool
  • Circulation Symmetry preserving One-Arm mode
  • Circulation Symmetry preserving Nexus / ITD “sandwich” mode


Dynamic FCoE

J. Metz defined this functionality fairly nicely again in 2014 (love the diagrams!), so I’ll refer you to his weblog for particulars and the diagrams.

The fundamental concept is computerized FCoE over a FabricPath material, unifying the LAN and SAN infrastructures. No further licensing prices.

The cool half is that the automation kicks in when FCoE is configured on the leaf swap. The phrase “dynamic” refers to:

  • NX-OS routinely creates a full mesh of digital hyperlinks between leaf switches
  • NX-OS routinely creates the digital Fibre Channel (vFC) ports
  • NX-OS routinely creates the VE_Port kind and the ISL

I ought to in all probability observe, that is “new” expertise (as of 3-Four years in the past) and would possibly actually disturb your SAN group if you happen to suggest utilizing it. In that case, be ready for phrases like “unproven”, “dangerous”, and many others. Don’t most SAN individuals typically assume Networking persons are wild and loopy danger takers?

My solely query: When can we’ve got this for VXLAN? Yeah, okay, the Nexus 9K’s want full FCF performance too, and that in all probability comes first.



I haven’t used ITD within the discipline but, nor Dynamic FCoE, so in case you have, and have one thing you’d care to go alongside (good, dangerous, ugly), please accomplish that by way of a Remark!

Feedback are welcome, each in settlement or constructive disagreement in regards to the above. I get pleasure from listening to from readers and carrying on deeper dialogue by way of feedback. Thanks upfront!


Hashtags: #CiscoChampion #CiscoNexus

Twitter: @pjwelcher

This text initially appeared on the Netcraftsmen weblog.

Comments are closed.