I am using a 3825 as an edge router. It has only the two integrated gig ports hard set to 100/full. On my side is a 6509 (hard set to 100/full)and on the ISP side is a Riverstone 10/100 switch (set to 100/full). I can only drive the throughput to about 2000pps at about 10Mbs (combined both in/out). Once in a while I see spikes over 10Mbs.
I noticed that my 6509 doesn't support flow control on the 100BaseT ports but the 3825 says flow control is set to XON. I can't find any command to change this. Could this be a problem?
Cisco 6509 switch if the switch running with CatOS then you can do the flow control by changing the port speed.
??? I am running hybrid os but the "set port speed" command has no options for flow control that I can see.
When I do a "set port flowcontrol" command I get "flowcontrol mode 'on' not supported on module"
How are you testing pps? What size packets are you using?
Testing throughput or bandwidth using any kind of TCP/IP transfer will not be accuarate due to varying packet sizes, acknowledgements, and other TCP overhead.
PPS is just by the counters on the interface. I am not tracking sizes. Bandwidth utilization is monitored by MRTG and flattens out at 10Mbs with spikes up to about 18Mbs.
ip address xxx.xxx.xxx 255.255.255.224
no ip redirects
no ip unreachables
no ip proxy-arp
ip nat inside
ip route-cache flow
no cdp enable
no mop enabled
hold-queue 2500 in
hold-queue 200 out
The config. looks fine. Did you confirm the ISP did not limit the throughput and find any packet when it reach 10M or 18Mbps ? Please provide the "sh int".
I am working with the ISP right now. I am just trying to rule out anything on my side. I noticed the flow control thing (see my first post) and didn't know if that would affect throughput.
Thanks for all the responses!
GigabitEthernet0/0 is up, line protocol is up
Hardware is BCM1125 Internal MAC, address is 0012.7f29.04c0 (bia 0012.7f29.04
Internet address is xxx.xxx.xxx.xxx/27
MTU 1500 bytes, BW 100000 Kbit, DLY 100 usec,
reliability 255/255, txload 11/255, rxload 2/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 100Mb/s, link type is autonegotiation, media type is RJ45
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 2w4d
Input queue: 0/2500/0/86 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/200 (size/max)
30 second input rate 1078000 bits/sec, 502 packets/sec
30 second output rate 4419000 bits/sec, 612 packets/sec
553006601 packets input, 3478560860 bytes, 0 no buffer
Received 0 broadcasts, 0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 watchdog, 940490 multicast, 0 pause input
0 input packets with dribble condition detected
690657734 packets output, 1645866688 bytes, 0 underruns
0 output errors, 0 collisions, 0 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
I don't find any problem of the interface. There is only one issue that the config. show it is manual config. of speed/duplex but it is auto from the sh int. Can you clarify it is manual setup ?
There is no packet drop at the interface, will the limitation is carried on the end device, e.g. your LAN side or the remote side behind the ISP ?
I suggest to disconnect the GE from the ISP and try to connect the GE to a PC then make a throughput test locally. If there is no problem, it is the ISP issue or the remote end host issue.
Hope this helps.