Duplicate packets b/w nodes

Whenever I ping between nodes in my infrastructure network, not to the
router but between wifi clients I get duplicate packets:

PING 192.168.0.149 (192.168.0.149): 56 data bytes
64 bytes from 192.168.0.149: icmp_seq=0 ttl=64 time=4.9 ms
64 bytes from 192.168.0.149: icmp_seq=0 ttl=64 time=6.1 ms (DUP!)
64 bytes from 192.168.0.149: icmp_seq=1 ttl=64 time=2.5 ms
64 bytes from 192.168.0.149: icmp_seq=1 ttl=64 time=3.6 ms (DUP!)
64 bytes from 192.168.0.149: icmp_seq=2 ttl=64 time=2.1 ms
64 bytes from 192.168.0.149: icmp_seq=2 ttl=64 time=3.2 ms (DUP!)
64 bytes from 192.168.0.149: icmp_seq=3 ttl=64 time=2.1 ms
64 bytes from 192.168.0.149: icmp_seq=3 ttl=64 time=3.2 ms (DUP!)
64 bytes from 192.168.0.149: icmp_seq=4 ttl=64 time=3.1 ms
64 bytes from 192.168.0.149: icmp_seq=4 ttl=64 time=4.2 ms (DUP!)

I don't get this when pinging wired connections even when between a
wired and wireless connection as well as the router/ap. What could the
possible issues be that would cause this?

Hardware:
SMC2804WBR(latest firmware)
IPW2100(latest linux os driver: http://ipw2100.sf.net)
DWL520+(latest linux os driver: http://acx100.sf.net)

Thanks for the help.

On Mon, 12 Jul 2004 04:16:02 GMT, Elliot English
<(E-Mail Removed)> wrote:

>Whenever I ping between nodes in my infrastructure network, not to the
>router but between wifi clients I get duplicate packets:
>
>PING 192.168.0.149 (192.168.0.149): 56 data bytes
>64 bytes from 192.168.0.149: icmp_seq=0 ttl=64 time=4.9 ms
>64 bytes from 192.168.0.149: icmp_seq=0 ttl=64 time=6.1 ms (DUP!)
>64 bytes from 192.168.0.149: icmp_seq=1 ttl=64 time=2.5 ms
>64 bytes from 192.168.0.149: icmp_seq=1 ttl=64 time=3.6 ms (DUP!)
>64 bytes from 192.168.0.149: icmp_seq=2 ttl=64 time=2.1 ms
>64 bytes from 192.168.0.149: icmp_seq=2 ttl=64 time=3.2 ms (DUP!)
>64 bytes from 192.168.0.149: icmp_seq=3 ttl=64 time=2.1 ms
>64 bytes from 192.168.0.149: icmp_seq=3 ttl=64 time=3.2 ms (DUP!)
>64 bytes from 192.168.0.149: icmp_seq=4 ttl=64 time=3.1 ms
>64 bytes from 192.168.0.149: icmp_seq=4 ttl=64 time=4.2 ms (DUP!)
>
>I don't get this when pinging wired connections even when between a
>wired and wireless connection as well as the router/ap. What could the
>possible issues be that would cause this?

See:
http://www.island.net/~andyb/#Duplicate
http://www.mail-archive.com/isp-linu.../msg02871.html
for some clues. I don't really understand the 2nd message, but
perhaps it might be useful.

Notice that each dupe adds 50% more latency to the ping. Last time I
saw this was when the cts/rts flow control threshold was screwed up in
the access point firmware. The access point is suppose to send and
receive packets smaller than the threshold without any flow control.
The threshold was suppose to be negotiated with the client radios to
insure that everyone's threshold was identical. This didn't work.
I'll spare you the details (unless you want them), but the result was
that small packets were always sent twice. Nobody noticed because it
didn't really affect thruput very much. Anything above the threshold
would get sent only once. Try sending larger ping packets and see if
the effect goes away. If it does, start looking for new access point
firmware.

You can also create the same effect by running a sniffer on the same
box as the firewall. I ran into this when running snort on a hacked
version of the WRT54G firmware. See:
http://www.oreillynet.com/cs/user/view/wlg/2086
for some clues (but no real answers).

I've also seen chronic dupes when there is substantial interference.
The latency is always an integral multiple of the normal ping time
resulting from packet loss and a retrans. This is not the case in
your situation, but I thought I would throw it in for reference.

>Hardware:
>SMC2804WBR(latest firmware)
>IPW2100(latest linux os driver: http://ipw2100.sf.net)
>DWL520+(latest linux os driver: http://acx100.sf.net)
>
>Thanks for the help.


--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

Jeff Liebermann wrote:
> On Mon, 12 Jul 2004 04:16:02 GMT, Elliot English
> <(E-Mail Removed)> wrote:
>
>
>>Whenever I ping between nodes in my infrastructure network, not to the
>>router but between wifi clients I get duplicate packets:
>>
>>PING 192.168.0.149 (192.168.0.149): 56 data bytes
>>64 bytes from 192.168.0.149: icmp_seq=0 ttl=64 time=4.9 ms
>>64 bytes from 192.168.0.149: icmp_seq=0 ttl=64 time=6.1 ms (DUP!)
>>64 bytes from 192.168.0.149: icmp_seq=1 ttl=64 time=2.5 ms
>>64 bytes from 192.168.0.149: icmp_seq=1 ttl=64 time=3.6 ms (DUP!)
>>64 bytes from 192.168.0.149: icmp_seq=2 ttl=64 time=2.1 ms
>>64 bytes from 192.168.0.149: icmp_seq=2 ttl=64 time=3.2 ms (DUP!)
>>64 bytes from 192.168.0.149: icmp_seq=3 ttl=64 time=2.1 ms
>>64 bytes from 192.168.0.149: icmp_seq=3 ttl=64 time=3.2 ms (DUP!)
>>64 bytes from 192.168.0.149: icmp_seq=4 ttl=64 time=3.1 ms
>>64 bytes from 192.168.0.149: icmp_seq=4 ttl=64 time=4.2 ms (DUP!)
>>
>>I don't get this when pinging wired connections even when between a
>>wired and wireless connection as well as the router/ap. What could the
>>possible issues be that would cause this?
>
>
> See:
> http://www.island.net/~andyb/#Duplicate
> http://www.mail-archive.com/isp-linu.../msg02871.html
> for some clues. I don't really understand the 2nd message, but
> perhaps it might be useful.
>
> Notice that each dupe adds 50% more latency to the ping. Last time I
> saw this was when the cts/rts flow control threshold was screwed up in
> the access point firmware. The access point is suppose to send and
> receive packets smaller than the threshold without any flow control.
> The threshold was suppose to be negotiated with the client radios to
> insure that everyone's threshold was identical. This didn't work.
> I'll spare you the details (unless you want them), but the result was
> that small packets were always sent twice. Nobody noticed because it
> didn't really affect thruput very much. Anything above the threshold
> would get sent only once. Try sending larger ping packets and see if
> the effect goes away. If it does, start looking for new access point
> firmware.
>
> You can also create the same effect by running a sniffer on the same
> box as the firewall. I ran into this when running snort on a hacked
> version of the WRT54G firmware. See:
> http://www.oreillynet.com/cs/user/view/wlg/2086
> for some clues (but no real answers).
>
> I've also seen chronic dupes when there is substantial interference.
> The latency is always an integral multiple of the normal ping time
> resulting from packet loss and a retrans. This is not the case in
> your situation, but I thought I would throw it in for reference.
>
>
>>Hardware:
>>SMC2804WBR(latest firmware)
>>IPW2100(latest linux os driver: http://ipw2100.sf.net)
>>DWL520+(latest linux os driver: http://acx100.sf.net)
>>
>>Thanks for the help.
>
>
>
I tried pinging with a larger packet, at 1k the dups weren't as
regularly place but they were all there. At 4k they were completely gone
as you had suggested. What can I do to fix this if there is no alternate
firmware for my router/ap and I only have one ap so it's not that two
are on the channel. I am certain that interference isn't an issue.

Thanks.