crossover cables / OpenBSD crashes

Seth Gordon sgordon at kenan.com
Mon Jan 29 12:19:22 EST 2001


My home network used to look like this:

                    ---------------
   --------------   |W  Linksys  L|   -------   -----------
---| DSL router |---|A  firewall A|---# HUB |---| petunia |
   --------------   |N    box    N|   -------   -----------
                    ---------------   | |   |
                                      | |   |
                         ----------------  ------------
                         | horse-nettle |  | eggplant |
                         ----------------  ------------

(The # represents the "uplink" port on the hub.  Horse-nettle has two
links to the hub because it has two Ethernet cards -- before I got the
Linksys as a housewarming present, I was planning to configure
horse-nettle as a firewall.)

Unfortunately, when I try to run a DNS server on horse-nettle,
machines outside the firewall can't access it.  No problem, I think --
I'll just rearrange the network so it looks like this:

                                       --------------
   --------------   ----------------   |W Linksys  L|   -------
---| DSL router |---| horse-nettle |---|A firewall A|---# HUB |
   --------------   ----------------   |N   box    N|   -------
                                       --------------   |     |
                                              -----------     |   
                                              | petunia |     |
                                              -----------     |
                                                   ------------
                                                   | eggplant |
                                                   ------------

QUESTION THE FIRST: In this configuration, which connections, if any,
require crossover cables?  (In the previous configuration, nothing had
crossover cables.  The Linksys has a button you can push depending on
whether you're connecting its LAN port to a hub or directly to another
machine.)

QUESTION THE SECOND: After setting things up like this, horse-nettle
(which runs OpenBSD 2.8) suffers a kernel crash a few minutes after
going into multi-user mode:

    uvm_fault(0xe3f180c8, 0x0, 0, 1) -> 1
    Kernel: page fault trap, code=0
    Stopped at _ipv4_input+0x1eb: movl 0x8(%edi), %esi

I've tried changing the IP addresses on the Ethernet cards (which
needed changing anyway), unplugging one card from the router, plugging
that card back into the router and unplugging the other card from the
firewall.  (I now have it unplugged at both sides, but I don't know
yet whether or not that solves the problem, because I got sick of
watching the machine go up and down and went to bed.)  At first, I
thought that the OS was crashing because of the card configuration. 
Then, I thought it was because of improper cabling.  Now, I think /bsd
file is corrupt (it's not like the motherboard and hard drive are
using state-of-the-art error-reduction technology).  What other
diagnoses are possible?  Aside from following the instructions in
crash(8) and ftp'ing over a fresh 2.8 kernel, what should I do?

-- 
"The big dig might come in handy ... for a few project managers
 whom I think would make great landfill."  --Elaine Ashton
== seth gordon == sgordon at kenan.com == standard disclaimer ==
== documentation group, kenan systems corp., cambridge, ma ==
-
Subcription/unsubscription/info requests: send e-mail with
"subscribe", "unsubscribe", or "info" on the first line of the
message body to discuss-request at blu.org (Subject line is ignored).



More information about the Discuss mailing list