Apparently you can cable the A-side Fabric Interconnect to the right IOM in a chassis, and it works just fine.

ucs_cross

You can even look at the DCE interfaces on a VIC in this chassis and see that the paths have been flipped:

ucs_cross2

This is not true for the “correctly” cabled chassis, where the A-side traces occupy the first two slots:

ucs_cross3

The first two interfaces will always go to the left IOM because the backplane traces are cabled that way. However, the IOM is aware of what FI it’s connected to:

ucs_cross4

The DCE interfaces receive this information from the IOM they connect to, and as a result, the VIC knows to pin A-side vNICs to the right traces. We’ve “crossed the streams” as it were, but nothing north of the Fabric Interconnects would know this.

So while there’s certainly an argument for cabling things straight-down to lessen any confusion, it’s interesting to see that there’s absolutely no such thing as an A-side or B-side IOM or IOM slot. Both IOMs are identical, and the slots are treated the same way. They just assume the fabric identity of the FI they’re connected to.

I’m up late staging some UCS gear and just wanted to share.


Matt Oswalt

Matt Oswalt is an all-around technology nerd, currently focusing on networking, software development, and everything in between. He is at his happiest in front of a keyboard, next to a brewing kettle, or wielding his silo-smashing sledgehammer. He spends his days diving deep into the intersection of networking and software, and likes to blog about his experiences when he comes up for air. You can follow him on Twitter or LinkedIN.