[c-nsp] overruns on a lightly loaded ethernet port
Robert Larsen
rkml67 at gmail.com
Thu May 12 16:29:45 EDT 2005
In my experience, this could still be enough to cause the overruns. It
really depends on the length/size of the traffic burst, packet size, etc.
Rob.
-----Original Message-----
From: Burton Windle [mailto:bwindle at fint.org]
Sent: 12 May 2005 03:22
To: Robert Larsen
Subject: RE: [c-nsp] overruns on a lightly loaded ethernet port
This 3640 has 8 T1s connected to it, so it could (at max) only see 12.3
megabits going in or out of that FE, which I think would eliminate the
possibility of anything near line rate traversing it.
On Thu, 12 May 2005, Robert Larsen wrote:
> We've seen similar problems on FE interfaces on 2621 routers, and this is
> most likely because the burst of traffic is too much for the input buffer
on
> the interface to deal with, so the router has to effectively tail drop
> packets which causes the overruns in the interface stats.
>
> Although these "low-end" routers (2600 and 3600 series) have FE
interfaces,
> they have fairly limited buffers and processing power, and as such could
> never deal with traffic arriving at anything approaching line rate for any
> length of time.
>
> We've done some testing on a 2621XM a few years ago, and up to approx.
2Mbps
> with a packet size of 192 bytes (or greater) it just about coped (CPU was
> running at 80-100%), but any increase in traffic rate *or* decrease in
> packet size just maxed the box out completely.
>
> Cheers,
>
> Rob.
>
> -----Original Message-----
> From: cisco-nsp-bounces at puck.nether.net
> [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Burton Windle
> Sent: 11 May 2005 22:38
> To: cisco-nsp at puck.nether.net
> Subject: [c-nsp] overruns on a lightly loaded ethernet port
>
> I have a 3640, running 12.2(28), connected directly to a 6509 via 100bT.
> Every now and then, I see a large burst of 'overruns' reported on the
> 3640's FastEth int; however, this connection is very lightly used
> (~2megabit/sec average). I am under the impression that overruns on an
> Ethernet interface happen when one side tries to send too fast, but this
> interface never pushes more than 5 megabit/sec.
>
> 5-3640-WAN>sh proc cpu
> CPU utilization for five seconds: 17%/16%; one minute: 12%; five minutes:
> 17%
>
> !
> interface FastEthernet0/0
> description ***6509_PORT2/1***
> ip address 192.168.1.4 255.255.255.0
> ip route-cache flow
> speed 100
> full-duplex
>
> 5-3640-WAN>sh int fastEthernet 0/0
> FastEthernet0/0 is up, line protocol is up
> Hardware is AmdFE, address is 0006.2889.e301 (bia 0006.2889.e301)
> Description: ***5-6509-1_PORT2/1***
> Internet address is 192.168.1.4/24
> MTU 1500 bytes, BW 100000 Kbit, DLY 100 usec,
> reliability 255/255, txload 3/255, rxload 3/255
> Encapsulation ARPA, loopback not set
> Keepalive set (10 sec)
> Full-duplex, 100Mb/s, 100BaseTX/FX
> 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 2d01h
> Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
> Queueing strategy: fifo
> Output queue: 0/40 (size/max)
> 5 minute input rate 1304000 bits/sec, 592 packets/sec
> 5 minute output rate 1476000 bits/sec, 598 packets/sec
> 62322869 packets input, 4032355232 bytes
> Received 157135 broadcasts, 0 runts, 0 giants, 0 throttles
> 750 input errors, 0 CRC, 0 frame, 750 overrun, 0 ignored
> 0 watchdog
> 0 input packets with dribble condition detected
> 60861002 packets output, 108203385 bytes, 0 underruns
> 0 output errors, 0 collisions, 0 interface resets
> 0 babbles, 0 late collision, 0 deferred
> 0 lost carrier, 0 no carrier
>
>
>
> 5-6509-1> sh port 2/1
> Port Name Status Vlan Duplex Speed Type
> ----- -------------------- ---------- ---------- ------ ----- ------------
> 2/1 ***5-3640-WAN*** connected 1 full 100 10/100BaseTX
>
> Port AuxiliaryVlan AuxVlan-Status InlinePowered PowerAllocated
> Admin Oper Detected mWatt mA @42V
> ----- ------------- -------------- ----- ------ -------- ----- --------
> 2/1 none none off off no 0 0
>
>
> Port Security Violation Shutdown-Time Age-Time Max-Addr Trap IfIndex
> ----- -------- --------- ------------- -------- -------- -------- -------
> 2/1 disabled shutdown 0 0 1 disabled 59
>
> Port Num-Addr Secure-Src-Addr Age-Left Last-Src-Addr
> Shutdown/Time-Left
> ----- -------- ----------------- -------- -----------------
> ------------------
> 2/1 0 - - - -
> -
>
> Port Broadcast-Limit Multicast Unicast Total-Drop
> -------- --------------- --------- ------- --------------------
> 2/1 - - - 0
>
> Port Send FlowControl Receive FlowControl RxPause TxPause
> admin oper admin oper
> ----- -------- -------- --------- --------- ---------- ----------
> 2/1 off off off off 0 0
>
> Port Status Channel Admin Ch
> Mode Group Id
> ----- ---------- -------------------- ----- -----
> 2/1 connected off 4 0
>
> Port Align-Err FCS-Err Xmit-Err Rcv-Err UnderSize
> ----- ---------- ---------- ---------- ---------- ---------
> 2/1 0 0 0 0 0
>
> Port Single-Col Multi-Coll Late-Coll Excess-Col Carri-Sen Runts
Giants
> ----- ---------- ---------- ---------- ---------- --------- ---------
> ---------
> 2/1 0 0 0 0 0 0
> 0
>
> Port Last-Time-Cleared
> ----- --------------------------
> 2/1 Fri Apr 1 2005, 22:16:01
>
>
>
> Any idea what would be causing these?
>
> --
> Burton Windle bwindle at fint.org
>
> _______________________________________________
> 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/
>
> _______________________________________________
> 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/
>
--
Burton Windle bwindle at fint.org
More information about the cisco-nsp
mailing list