[PLUG] Random? Crashing of router

[Aaron Ten Clay]

I'm having some various issues with a system I'm using as a router,
and am hoping for some general advice.

I think part of the problem may be two Compaq dual port Nettelligent
ethernet cards using the thunderlan (TLAN) driver. These cards have
had problems in two seperate systems, running both 2.4 and 2.6
kernels.

When I awoke this morning, no network traffic could go anywhere. I
investigated and found the router box completely locked up. No
messages had been written to console, and upon hard-rebooting and
checking the syslog, I found a bunch of garbage. Here's a snippit:

--- snip ---
Jan 20 03:58:24 router NETDEV WATCHDOG: eth1: transmit timed out
Jan 20 03:58:24 router TLAN: eth1: Starting autonegotiation.
Jan 20 03:58:26 router TLAN: eth1: Autonegotiation complete.
Jan 20 03:58:26 router TLAN: eth1: Link active with AutoNegotiation
enabled, at 100Mbps Full-Duplex
Jan 20 03:58:26 router TLAN: Partner capability: 10BaseT-HD 10BaseT-FD
100baseTx-HD 100baseTx-FD<NULL>
Jan 20 03:58:46 router NETDEV WATCHDOG: eth1: transmit timed out
Jan 20 03:58:46 router TLAN: eth1: Starting autonegotiation.
Jan 20 03:58:48 router TLAN: eth1: Autonegotiation complete.
Jan 20 03:58:48 router TLAN: eth1: Link active with AutoNegotiation
enabled, at 100Mbps Full-Duplex
Jan 20 03:58:48 router TLAN: Partner capability: 10BaseT-HD 10BaseT-FD
100baseTx-HD 100baseTx-FD<NULL>
^@^@^@^@^@^@^@^@^@^@^B^@^@^@X,t^@^P^B^@^@<FF><FF><FF><FF>^@^@^@^@^@^@^@^@
^@^@^@P^@^@^@^@^@^@^@^K^@^@^@ZipDrive100^@^@^@^@^@<E0><FF><FF><FF>vk^E^@P^@^@^@`4t^@^G^@^@^@^A^@^@^@Icons^@^@^@<F0><FF><FF><FF>8,t^@<B8>4t^@<^@<B4><E9><98><FF><FF><FF>nk
^@<AE><8E><99><B8><E4><AC>
<C4>^A^@^@^@^@<B8>^@^@^@^@^@^@^@^@^@<FF><FF><FF><FF><FF><FF><FF><FF>^C^@^@^@^X't^@^P^B^@^@<FF><FF><FF><FF>^@^@^@^@^@^@^@^@^V^@^@^@P^@^@^@^@^@^@^@^S^@^@^@PortableAudioPlayer^@^@^@^@^@<E0><FF><FF><FF>vk^E^@P^@^@^@<F0>,t^@^G^@^@^@^A^@^@^@Icons^@^@^@<A8><FF><FF><FF>%^@S^@y^@s^@t^@e^@m^@R^@o^@o^@t^@%^@\^@s^@y^@s^@t^@e^@m^@3^@2^@\^@s^@h^@e^@l^@l^@3^@2^@.^@d^@l^@l^@,^@-^@2^@9^@9^@^@^@^@^@^@^@^@^@<D8><FF><FF><FF>vk^K^@^V^@^@^@p-t^@^A^@^@^@^A^@^@^@NoSoftEject^@^@^@
^@^@<E0><FF><FF><FF>0^@x^@0^@0^@0^@0^@0^@0^@0^@1^@^@^@^@^@^@^@^@^@<A8><FF><FF><FF>nk
^<F1><9B><B8><E4><AC><C4>^A^@^@^@^@<B8>^@^@^@^@^@^@^@^@^@<FF><FF><FF><FF><FF><FF><FF><FF>^B^@^@^@<98>.t^@^P^B^@^@<FF><FF><FF><FF>^@^@^@^@^@^@^@^@
^@^@^@N^@^@^@^@^@^@^@^G^@^@^@Printer^@<E0><FF><FF><FF>vk^E^@N^@^@^.t^@^G^@^@^@^A^@^@^@Icons^@^@^@<A8><FF><FF><FF>%^@S^@y^@s^@t^@e^@m^@R^@o^@o^@t^@%^@\^@s^@y^@s^@t^@e^@m^@3^@2^@\^@s^@h^@e^@l^@l^@3^@
2^@.^@d^@l^@l^@,^@-^@1^@7^@^@^@^@^@^@^@^@^@^@^@<E0><FF><FF><FF>vk^E^@^P^@^@^@<80>.t^@^A^@^@^@^A^@^@^@Label^@^@^@<E8><FF><FF><FF>P^@r^@i^@n^@t^@e^@r^@^@^@^@^@^@^@<F0><FF><FF><FF><E8>-t^@`.t^@h^@
<A8><FF><FF><FF>nk
^<F1><9B><B8><E4><AC><C4>^A^@^@^@^@<B8>^@^@^@^@^@^@^@^@^@<FF><FF><FF><FF><FF><FF><FF><FF>^B^@^@^@<B0>/t^@^P^B^@^@<FF><FF><FF><FF>^@^@^@^@^@^@^@^@
^@^@^@P^@^@^@^@^@^@^@^G^@^@^@Scanner^@<E0><FF>
Jan 20 07:35:34 router syslog-ng[1130]: syslog-ng version 1.6.5 starting
--- snip ---

Note the series of messages from NETDEV WATCHDOG to ethX: Link active
occurs several hundred times in the logs prior to the end of logging.
I am guessing the garbage is due to fs corruption after the freeze?

The only conclusions I can draw are related to the TLAN drivers and/or
hardware, since this box ran fine with 2 Realtek gigabit NICs and two
Intel pro100 NICs for months. The more I think about it, all the
problems related to this box have affected networks connected via TLAN
cards. The only time other networks lost connectivity to the router
was today when it locked up completely.

Any advice would be appreciated. Google found some discussions about
TLAN driver issues, especially on other platforms, but nothing seemed
relevant to this. (Maybe I don't know how to search effectively... I
tried "linux (thunderlan OR tlan) problems"

I'm running kernel 2.4.28-gentoo-r5 on this box, patched with
2.4.28-ja2.diff from http://www.ssi.bg/~ja/#combinations

TIA,
Aaron