10-02-2014 04:53 PM - edited 03-01-2019 11:51 AM
I am putting together a UCS runbook for an operational team to use to help with troubleshooting Cisco UCS. I was leveraging some Cisco Live presentations including 'BRKCOM-3002 - UCS Performance Troubleshooting'. In this presentation, when looking at congestion inside UCS the presenter makes a reference to check for pause frames...in this customer's environment I ran the same command for documenting the procedure and noticed many Tx pause frames. Is this normal? Unfortunately I have never checked these counters in other environments so I have nothing to base it off. There are 0 pause frames on LAN uplink interfaces and only a few 100 pause frames on the server downlink interfaces to the FEX. This environment is running VMware ESXi 5.5 U1.
ucs-A(nxos)# show interface ethernet 1/1/1 priority-flow-control
============================================================
Port Mode Oper(VL bmap) RxPPP TxPPP
============================================================
Ethernet1/1/1 Auto On (18) 0 607998
As you can see there are many pause frames being sent from the blades...
Ethernet1/1/1 Auto On (18) 0 607998
Ethernet1/1/2 Auto Off 0 0
Ethernet1/1/3 Auto On (18) 0 647302
Ethernet1/1/4 Auto Off 0 0
Ethernet1/1/5 Auto On (18) 0 686730
Ethernet1/1/6 Auto Off 0 0
Ethernet1/1/7 Auto On (18) 0 613622
Ethernet1/1/8 Auto Off 0 0
Ethernet1/1/9 Auto On (18) 0 447698
Ethernet1/1/10 Auto Off 0 0
Ethernet1/1/11 Auto On (18) 0 479446
Ethernet1/1/12 Auto Off 0 0
Ethernet1/1/13 Auto On (18) 0 429074
Ethernet1/1/14 Auto Off 0 0
Ethernet1/1/15 Auto On (18) 0 406480
And 0 pause frames on LAN uplinks Eth1/1-4 and not many on the server downlinks Eth1/9-16.
Ethernet1/1 Auto Off 0 0
Ethernet1/2 Auto Off 0 0
Ethernet1/3 Auto Off 0 0
Ethernet1/4 Auto Off 0 0
Ethernet1/5 Auto Off 0 0
Ethernet1/6 Auto Off 0 0
Ethernet1/7 Auto Off 0 0
Ethernet1/8 Auto Off 0 0
Ethernet1/9 Auto Off 96 414
Ethernet1/10 Auto Off 16 362
Ethernet1/11 Auto Off 24 410
Ethernet1/12 Auto Off 18 452
Ethernet1/13 Auto Off 26 318
Ethernet1/14 Auto Off 402 354
Ethernet1/15 Auto Off 876 308
Ethernet1/16 Auto Off 114 310
If I am pointing operational administrators to go check counters then I want to be clear what are expected returns. Any insight is appreciated.
Cheers,
Matt
10-06-2014 06:33 AM
Have a look at
http://www.cisco.com/c/en/us/td/docs/switches/datacenter/nexus5000/sw/troubleshooting/guide/N5K_Troubleshooting_Guide/n5K_ts_fcoe.html
I have seen this in multihop FCoE installations, specially with problems of FCoE disk arrays.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide