Nexus 5010 high Tx Pause

[kshah2spires]

I have a Server with Emulex OCE10102 CNA connected to Nexus 5010 I am seeing a very high "Tx Pause" on the Nexus side. Here is the output

sh int ethernet 1/8

Ethernet1/8 is up

  Hardware: 1000/10000 Ethernet, address: 000d.ecf5.bccf (bia 000d.ecf5.bccf)

  Description: *** AVAILABLE/SHUTDOWN ***

  MTU 1500 bytes, BW 10000000 Kbit, DLY 10 usec,

     reliability 255/255, txload 4/255, rxload 2/255

  Encapsulation ARPA

  Port mode is trunk

  full-duplex, 10 Gb/s, media type is 10g

  Beacon is turned off

  Input flow-control is on, output flow-control is on

  Rate mode is dedicated

  Switchport monitor is off

  Last link flapped 07:38:24

  Last clearing of "show interface" counters never

  30 seconds input rate 877536 bits/sec, 495 packets/sec

  30 seconds output rate 243574056 bits/sec, 14850 packets/sec

  Load-Interval #2: 5 minute (300 seconds)

    input rate 108.71 Mbps, 6.93 Kpps; output rate 183.84 Mbps, 11.68 Kpps

  RX

    812395033 unicast packets  113123 multicast packets  0 broadcast packets

    812508156 input packets  1542522706844 bytes

    724639332 jumbo packets  0 storm suppression packets

    0 runts  0 giants  0 CRC  0 no buffer

    0 input error  0 short frame  0 overrun   0 underrun  0 ignored

    0 watchdog  0 bad etype drop  0 bad proto drop  0 if down drop

    0 input with dribble  0 input discard

    0 Rx pause

  TX

    626577215 unicast packets  18338809 multicast packets  0 broadcast packets

    644916024 output packets  1066177460879 bytes

    498041692 jumbo packets

    0 output errors  0 collision  0 deferred  0 late collision

    0 lost carrier  0 no carrier  0 babble

   15648457 Tx pause

  15 interface resets

Can someone suggest what could be the cause of high "Tx Pause". This is probably slowing down my server signficantly. I have updated the CNA with latest firmware

[ddulek]

http://www.cisco.com/en/US/docs/switches/datacenter/nexus5000/sw/troubleshooting/guide/n5K_ts_fcoe.htm

In the PCF section, it appears to be a flow control issue on that port on the switch.  I have a similar issue on a 5020 and have a TAC open.  It is "supposed" to pause when there is congestion but I do not think this server could be pushing 4 10G FCoE ports since the server it connect to is 1G and the storage is 4G.  If this was not a production server, I would probably test this:

Use the following commands under "interface ethx/y" to enable link pause instead of PFC with DCBX capable devices:

• "no priority-flow-control mode on"

• "flowcontrol receive on"

• "flowcontrol send on"

[kshah2spires]

Hi David

Good to know I am not the only one around. Yes I had gone through the document and had made the recommended changes. No change still

[Vu Phan]

The Tx pause from N5K means N5K try to send traffic to the other end, but the other side could not process fast enough. Thus N5K has to hold traffic.

Check the host side. If flowcontrol on, make sure it is on on both sides.

[kshah2spires]

The understanding I have been given is that the Tx Pause is a pause requested by Cisco switch the client. The Rx Pause is a pause requested by the client.

[vdsudame]

In FCoE environment TX pause from N5k to the server might not be unexpected based on the environment. We will need trace between n5k switchport and CNA to determine what is going on.

Thanks, Vinayak

[jeffreybunch]

Did you ever find a resolution to this? I am working with my server folks and there is an indication of a possible bug with Emulex where there would be Frame Loss and Flow Control issues. I am wondering if this is the indicator. I have the same Emulex CNAs you mentioned connected to 2232 FEXs.