Interface Discards

Unanswered Question
Feb 16th, 2010
User Badges:

Hello,


I was looking for some help or advice or maybe a link to point me in the right direction. I"m having an issue with switch link going back to my router that is experiance a lot of transmit discards. I was hoping for some help into what causes this and how to troubleshoot the issue. I run into errors and issues all the time and sometimes just don't know where to start the troubleshoot process.


Thanks

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
tdistlists Tue, 02/16/2010 - 07:25
User Badges:

Hey,


According to cisco transmit discards / output discards typically refer to network congestion -- but in my experience they're often caused by faulty hardware.


First steps I would take would be to replace connections (fiber,cat5,etc...). I would also change gbics, sfps, and/or switchports on both the switch and the router -- whatever applies to your situation. Also, check the duplex is full.


If this doesn't help with your output discards I'd monitor the throughput on that interface. Is it attempting to exceed its maximum bandwidth? At a minimum you can change the load-interval to 30 seconds and watch the output rate. Or setup a span session for that link, and use tcpdump or wireshark to see what traffic is using that link.


This is a decent docs on common interface issues:  http://www.cisco.com/en/US/products/hw/switches/ps700/products_tech_note09186a008015bfd6.shtml


Let us know what you find.

ultimatep Thu, 09/10/2015 - 02:41
User Badges:

Hi.

Interface discards are not always a sign of faulty hardware.  In fact they are in a lot of configurations part of normal operation based on the configuration of each switch/port.  You will typically see discard rates increase and decrease on interfaces.  For example: you have two switches:

  switch 1 has vlans 10, 20 & 30 configured and uses int 0/48 as a 802.1q trunk port configured to allow vlans 10, 20, 30 to send/receive.

  switch 2 has only vlans 10 and 20 configured and uses int 0/48 as an 802.1q trunk port configured to allow vlans 10, 20 to send/receive.

  in this configuration you would expect to see discards because switch 1 int 0/48 will be sending vlan 30 traffic - but switch 2 int 0/48 is only configured for accepting vlans 10, 20 therefore it will drop/discard all the vlan 30 traffic because it has been configured to do just that.

The same can apply where both switch 1 and 2 have all the same vlan configurations for their 802.1q trunk ports, but int 0/1 thru int 0/24 on switch 2 are configured on a static vlan of 10 then it would discard any traffic that hits these interfaces that is tagged for vlan 20 or 30.

I would check all your configurations before jumping to the conclusion that you have faulty hardware.

Let me know what you find?

Actions

This Discussion