[c-nsp] Pause Output on interface...

Blake Dunlap ikiris at gmail.com
Fri Oct 12 15:26:52 EDT 2012


This is flow control. It is a signaling method of an upstream device to
indicate that it cannot service more input data because it's ability to
drain it's input queue is temporarily congested. Generally this is due to
an oversubscribed backplane being full, either the primary backplane of the
device, or the uplink to the backplane that the port shares with other
ports in an oversubscribed manner.

If you are seeing inputs, it means the device is recieving the pauses, if
you are seeing outputs, it means the device is generating them.

One of the main downsides of flow control is that it is for the entire
port, which can cause issues for other traffic unrelated to congested
networks.

Most people disable flow control the first time there is an issue, unless
they specifically want it and understand its issues with it, and suffer the
dropped packets on input, as opposed to stopping all output traffic on the
previous device, which could cause a cascade backwards on the network. That
being said, you still want to fix the underlying issue, which is causing
the pauses to be sent in the first place.

-Blake

On Fri, Oct 12, 2012 at 2:05 PM, Jonathan Charles <jonvoip at gmail.com> wrote:

> We have a Cisco 2921 running 15.2(3T);
>
> We are seeing PAUSE OUTPUT incrementing on the device, and every time it
> does, our users experience pretty nasty jitter and silence on VOIP calls.
>
> The link is 1000BaseLX with a 100MB committed rate; we rarely see more than
> 10-15MB of output or input traffic on the interface.
>
> Two questions:
>
> 1 - Who is originating the Pause Outputs? Cisco documentation is silent on
> the subject and TAC has given us conflicting answers.
> 2 - If generated locally, how can we stop them from doing so? The provider
> is saying these are causing the issues we are experiencing.
>
>
> ROUTER#sh int gi0/1
> GigabitEthernet0/1 is up, line protocol is up
>   Hardware is CN Gigabit Ethernet, address is 2894.0fbd.0091 (bia
> 2894.0fbd.0091)
>   MTU 1500 bytes, BW 100000 Kbit/sec, DLY 10 usec,
>      reliability 255/255, txload 13/255, rxload 17/255
>   Encapsulation 802.1Q Virtual LAN, Vlan ID  1., loopback not set
>   Keepalive set (10 sec)
>   Full Duplex, 1Gbps, media type is LX
>   output flow-control is XON, input flow-control is XON
>   ARP type: ARPA, ARP Timeout 04:00:00
>   Last input 00:00:00, output 00:00:00, output hang never
>   Last clearing of "show interface" counters 16:53:34
>   Input queue: 1/4096/0/1204 (size/max/drops/flushes); Total output drops:
> 0
>   Queueing strategy: Class-based queueing
>   Output queue: 0/1000/0 (size/max total/drops)
>   30 second input rate 6684000 bits/sec, 3871 packets/sec
>   30 second output rate 5105000 bits/sec, 2906 packets/sec
>      87183773 packets input, 3179860385 bytes, 0 no buffer
>      Received 0 broadcasts (0 IP multicasts)
>      0 runts, 0 giants, 0 throttles
>      0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
>      0 watchdog, 0 multicast, 0 pause input
>      60076680 packets output, 4290590647 bytes, 0 underruns
>      0 output errors, 0 collisions, 0 interface resets
>      0 unknown protocol drops
>      0 babbles, 0 late collision, 0 deferred
>      0 lost carrier, 0 no carrier, 1252219 pause output
>      0 output buffer failures, 0 output buffers swapped out
> ROUTER#
>
> Thanks!
>
>
>
> Jonathan
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>


More information about the cisco-nsp mailing list