For e.g downloading a file of 37.8 MB in a LAN
through ftp from my machine to adjacent machine, my
"total_received_pkts" variable in netif_rx() function
in net/core/dev.c file, shows 13930 packets for this
particular file download. The total_received_pkts
variable in netif_rx() is incremented everytime when a
pkt is received by eth0. Now, tcpdump/tcptrace shows
me that total transferred packets are 38752 for this
particular file and maximum segment size (or pkt size)
is 1448 bytes. [The LAN has MTU of 1500 bytes].
So there is a great difference (in # pkts received by
the kernel) between what the kernel is showing me and
what the tcpdump/tcptrace shows. There is no
fragmentation involved on the LAN machines.
This makes my received bytes calculation too
wrong (because I am getting wrong number of received
pkts).
Could anybody please tell me where I am getting
it wrong when the tcpdump/tcptrace utility prints it
right ? I have repeated the experiment several times
(with udp, tcp protocol) but I couldnt get the reason.
I hope to get a reply.
Thank you very much for your precious time!
Gautam.
__________________________________________________
Do you Yahoo!?
Yahoo! Web Hosting - Let the expert host your site
http://webhosting.yahoo.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/