Re: How to survive in a Micro$oft environment??

From: Eric Brunet (ebrunet@clipper.ens.fr)
Date: Fri Feb 25 2000 - 17:51:54 EST


On 25 Feb 2000 20:31:39 GMT, Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
>> being returned to Linux is causing severe network problems, incluing
>> DHCP broadcast storms. Caldera's LIZARD utility gets into a state
>> whereby it starts flooding the network with DHCP broadcasts that will
>> crash local and remote networks (we have brought down our T1 with US
>
>Now thats interesting. I've seen a couple of Caldera reports but no others.
>Does Lizard do its own dhcp differently to the other dhcp packages ?
>
I have a similar problem here:
My ISP is putting some limitation on the traffic uploaded and claims that
from time to time my machine starts saturating the bandwith emitting
data. I have set up a firewall with ipchains which counts evrything that
goes, but it doesn't see this mysterious data. Other people from the
same ISP have reported the same problem, and it seems that only linux
computers generate that kind of bursts. I have not myself any tcpdump
data to show, but another user has mailed me this snippet:

20:47:56.069943 212.198.169.114.68 > 212.198.0.69.67: xid:0x69444112
                secs:3 C:212.198.169.114 [|bootp]
20:47:56.109949 212.198.169.114.68 > 212.198.0.69.67: xid:0x69444112
                secs:3 C:212.198.169.114 [|bootp]
20:47:56.149926 212.198.169.114.68 > 212.198.0.69.67: xid:0x69444112
                secs:3 C:212.198.169.114 [|bootp]
20:47:56.150106 212.198.169.114.68 > 212.198.0.69.67: xid:0x69444112
                secs:3 C:212.198.169.114 [|bootp]
20:47:56.189924 212.198.169.114.68 > 212.198.0.69.67: xid:0x69444112
                secs:3 C:212.198.169.114 [|bootp]

(the dhcp protocol is used to allocate IPs.)

That looks like the problem described...

Please remember that ipchains hasn't logged anything, so if there is a
bug it must in the kernel. Remember also that the tcpdump above is not
from my computer, that there wasn't any firewall logging output on that
computer and that it might very well be a completely unrelated problem
(for instance, a user land bug in pump or dhcpcd).

The problem seems to start (when it starts) at about the time where
dhcpcd should send a lease RENEWING.

My machine is a P90 running 2.2.1, the network card is a ne2000 PCI (but
with the ISA driver. While I am here, one question: I have an isa ne2000
and a pci ne2000 in the same computer. The isa card must be eth0. How do
I use the pci driver for eth1 ?) dhcpcd is 1.3.16 from an old developper
package from redhat.

Éric Brunet

----------------------------------------------------------------------

----- End forwarded message -----

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Tue Feb 29 2000 - 21:00:14 EST