cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
15534
Views
30
Helpful
16
Replies

3850 Output Queue Drops

Otaku78
Level 1
Level 1

Hi all, I'm experiencing an issue with a WS-C3850-24S switch where a small percentage of packets are dropping from the output queues in about half of the trunks. I've done some research and found that it can be caused by an ingress interface that offers much more bandwidth than the egress interface. I'm not entirely sure this is what's causing it to be honest.

Ingress on this switch is a 2 x 10Gb EtherChannel trunk. Firmware version is 03.06.06E .

I've also tried to use some of the fixes described in these forums in regards to changing the queue parameters but have been unsuccessful in making those changes. I would really appreciate it if someone could help me out!

 

 Trunk Interface stats:

 

 reliability 255/255, txload 4/255, rxload 1/255

 Encapsulation ARPA, loopback not set
  Keepalive not set
  Full-duplex, 1000Mb/s, link type is auto, media type is 1000BaseBX-10U SFP
  input flow-control is off, output flow-control is unsupported
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 00:00:18, output never, output hang never
  Last clearing of "show interface" counters 3w5d
  Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 959239
  Queueing strategy: fifo
  Output queue: 0/40 (size/max)
  30 second input rate 6068000 bits/sec, 2450 packets/sec
  30 second output rate 18567000 bits/sec, 1655 packets/sec
     147799652 packets input, 78574317037 bytes, 0 no buffer
     Received 405149 broadcasts (136421 multicasts)
     0 runts, 0 giants, 0 throttles
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     0 watchdog, 136421 multicast, 0 pause input
     0 input packets with dribble condition detected

 237563053 packets output, 252406070491 bytes, 0 underruns
 959239 output errors, 0 collisions, 0 interface resets

 

QoS Queue stats:

-------------------------------
Queue Buffers Enqueue-TH0 Enqueue-TH1 Enqueue-TH2
----- ------- ----------- ----------- -----------
    0       0           0           0  1192754861
    1       0           0           0   664465155
    2       0           0           0           0
    3       0           0           0           0
    4       0           0           0           0
    5       0           0           0           0
    6       0           0           0           0
    7       0           0           0           0
DATA Port:21 Drop Counters
-------------------------------
Queue Drop-TH0    Drop-TH1    Drop-TH2    SBufDrop    QebDrop
----- ----------- ----------- ----------- ----------- -----------
    0           0           0           0           0           0
    1           0           0     3619155           0           0
    2           0           0           0           0           0
    3           0           0           0           0           0
    4           0           0           0           0           0
    5           0           0           0           0           0
    6           0           0           0           0           0
    7           0           0           0           0           0
 AQM Broadcast Early WTD COUNTERS(In terms of Bytes)
--------------------------------------------------
  PORT TYPE          ENQUEUE             DROP
--------------------------------------------------
 UPLINK PORT-0        N/A               0
 UPLINK PORT-1        N/A               0
 UPLINK PORT-2        N/A               0
 UPLINK PORT-3        N/A               0
 NETWORK PORTS    21024980          140441674
 RCP PORTS               0                  0
 CPU PORT                0                  0
Note: Queuing stats are in bytes

16 Replies 16

Thank you again!

 

Jackie

Agree with Joseph. Jackie, you need to rule out congestion first to make sure it's not a bug. Check to see how much traffic is hitting your interfaces where the error counters are incrementing either by using SNMP or Wireshark, then clear the counters and monitor it for a while. If it's related to a bug the error counters will continue to increment even when there is little traffic and no congestion which is what happened to me. It my situation the errors occurred when the interfaces were at less than 2% Rx/Tx utilisation.


Review Cisco Networking products for a $25 gift card