The real benefit isn’t the crazy amount of bandwidth in the 2208XP. The Port-Channel feature is what really makes it!
The backplate speed from IOM to blades is x2 what the ports equal to FI’s. So:
2204 = 80Gb to blades.
2208 = 160Gb to blades.
This is good because newer 1280 Palo cards support 20Gb per FI uplink (The Cisco UCS M81KR do not support this). The 1280 has 4x10Gb lanes per channel (FI side). That sums up to 80Gb to FI! The new 2204 and 2208 IOM’s can handle that! There is a 1240 Palo out and it is half the lanes. So 2×10 per channel.
Compair Models
http://www.cisco.com/en/US/products/ps10277/prod_models_comparison.html
Once configured you can “connect nxos a” and see the actual Port-Channel’s
UCS-A(nxos)# show port-channel summary
Flags: D – Down P – Up in port-channel (members)
I – Individual H – Hot-standby (LACP only)
s – Suspended r – Module-removed
S – Switched R – Routed
U – Up (port-channel)
——————————————————————————–
Group Port- Type Protocol Member Ports
Channel
——————————————————————————–
1025 Po1025(SU) Eth NONE Eth1/1(P) Eth1/2(P)
1026 Po1026(SU) Eth NONE Eth1/3(P) Eth1/4(P)
1027 Po1027(SU) Eth NONE Eth1/5(P) Eth1/6(P)
1028 Po1028(SU) Eth NONE Eth1/7(P) Eth1/8(P)
1029 Po1029(SU) Eth NONE Eth1/9(P) Eth1/10(P)
1030 Po1030(SU) Eth NONE Eth1/11(P) Eth1/12(P)
The only configuration to use Port-Channles between FI’s and IOM’s is a global hardware policy set here
Also the Hybrid Display is improved in 2.0
Links to details below
http://www.cisco.com/en/US/prod/collateral/ps10265/ps10276/data_sheet_c78-675243.pdf
http://www.cisco.com/en/US/docs/unified_computing/ucs/hw/chassis/install/overview.html#wp1260436
Previous to these new IOM’s bandwidth was distributed over odd and even ports for odd and even numbers chassis slots. 40 Gbps was not truly available to any one server. Also had to uplink either with 1, 2, or 4 ports. See how the throughput flows below (pardon the very basic drawing).
Be First to Comment