Just trying to find out why I'm getting discards on a interface. Basically we have a remote office that connects to another office before connecting to the main backhaul.
SWITCH A ----> SWITCH B ----> SWICTH C
The Link between Switch A and B is 100Mbps
The Link between Switch B and C is 1000Mbps (1GB).
Looking into the history of the link between A and B we can see that majority of the time the traffic is at 60-65% but will spike upto 85% 1-2 times a week.
GigabitEthernet1/0/2 is up, line protocol is up (connected)
Hardware is Gigabit Ethernet, address is 001d.70b8.e282 (bia 001d.70b8.e282)
Description: to SWITCH-A TO SWITCH-B
MTU 1998 bytes, BW 100000 Kbit, DLY 100 usec,
reliability 255/255, txload 48/255, rxload 13/255
Encapsulation ARPA, loopback not set
Keepalive not set
Full-duplex, 100Mb/s, link type is auto, media type is 10/100/1000BaseTX SFP
input flow-control is off, output flow-control is unsupported
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 never
Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 1851383
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 5293000 bits/sec, 2799 packets/sec
5 minute output rate 19054000 bits/sec, 3739 packets/sec
91760131827 packets input, 21396599373912 bytes, 0 no buffer
Received 50352741 broadcasts (33957950 multicasts)
36 runts, 0 giants, 0 throttles
36 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 watchdog, 33957950 multicast, 0 pause input
0 input packets with dribble condition detected
123201990031 packets output, 112703016614681 bytes, 0 underruns
0 output errors, 0 collisions, 1 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier, 0 PAUSE output
0 output buffer failures, 0 output buffers swapped out
All switches are Cisco 3750. What would you recommend to resolve this issue as I've been advised that the QOS is fine and link will not be upgraded to a 1GB link?
Now looking into this I've read that either of the following can help:
- Increasing queue-limit
- Increasing hold queue
Cisco advising against increasing in hold queue as TCP timer can expire whilst the packet is still in the queue.
Anyone out there that can provide further input?