Question: bcm5700 and rx_discard stat

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi,

I'm getting a lot of dropped packets when I exceed about 400 Mbps (400 Mbps
in and 400 Mbps out) on a single network interface. This is a bcm5704-based
device. ethtool -S is showing me the rx_discard stat is increasing by
exactly the number of packets that have gone missing. My question is, what
exactly does this mean? Besides the obvious meaning that the packets were
dropped, why exactly is it dropping them? I assume there is some queue on
the nic itself that is getting full? I have tried changing the number of rx
descriptors up to the maximum value of 511, but to no avail (when I did this
with the tg3 driver, it locked up; with the bcm5700 driver, it took the
change but the behavior was the same).

More info - I am using the stock RedHat 2.4.21-20ELsmp kernel. I am using
iptables to NAT the packets before they are forwarded; they aren't being
delivered to userspace. The problem does not happen if I drop the data rate
down to about 350-360 Mbps. The problem does not happen if I use the 400
Mbps data rate but put the incoming stream on one interface and the outgoing
stream on another interface.

Thanks
Ed

-
: send the line "unsubscribe linux-net" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [Netdev]     [Ethernet Bridging]     [Linux 802.1Q VLAN]     [Linux Wireless]     [Kernel Newbies]     [Security]     [Linux for Hams]     [Netfilter]     [Git]     [Bugtraq]     [Yosemite News and Information]     [MIPS Linux]     [ARM Linux]     [Linux RAID]     [Linux PCI]     [Linux Admin]     [Samba]

  Powered by Linux