1.3 Development

For the past 3 days I've been trying to use m0n0wall on an old laptop, NEC 300mhz w/64mb ram, floppy drive to load smart boot manager, then booting off a 4gb hard drive.  I'm using 2 3com pcmcia nics and the 1.3b2 firmware as the 4.x BSD doesn't support this hardware.  Only 2 of my firewall settings are getting a lot of use on 2 separate ports.  I have a 15/2 connection, and the traffic graph confirmed that it can handle this speed, and the CPU graph appears fine.  I've removed the keyboard and setup a heatsink on the processor to try and rule out if its heat related, but the processor does not appear to be the cause

The pc locks up apparently randomly - anywhere between 8-12 hours.  The console shows nothing.

Is there a way I can log the pc so that I may find out what's causing the route to lock up?  I have MRTG setup, but that's only tracking outbound/inbound traffic.

Here is the only thing that shows up on the console when first booted, I'm not sure if it might indicate a problem elsewhere:

kernel: xl1: transmission error: 90
May 17 05:02:49    kernel: xl1: tx underrun, increasing tx start threshold to 180 bytes
May 17 05:02:49    kernel: xl1: transmission error: 90
May 17 05:02:49    kernel: xl1: tx underrun, increasing tx start threshold to 240 bytes
May 17 05:02:49    kernel: xl1: transmission error: 90
May 17 05:02:49    kernel: xl1: tx underrun, increasing tx start threshold to 300 bytes
May 17 05:02:50    kernel: xl0: transmission error: 90
May 17 05:02:50    kernel: xl0: tx underrun, increasing tx start threshold to 180 bytes
May 17 05:02:58    kernel: xl1: transmission error: 90
May 17 05:02:58    kernel: xl1: tx underrun, increasing tx start threshold to 360 bytes
May 17 05:03:33    kernel: xl0: transmission error: 90
May 17 05:03:33    kernel: xl0: tx underrun, increasing tx start threshold to 240 bytes
May 17 05:03:33    kernel: xl0: transmission error: 90
May 17 05:03:33    kernel: xl0: tx underrun, increasing tx start threshold to 300 bytes
May 17 05:03:41    kernel: xl0: transmission error: 90
May 17 05:03:41    kernel: xl0: tx underrun, increasing tx start threshold to 360 bytes
May 17 05:06:52    kernel: xl1: transmission error: 90
May 17 05:06:52    kernel: xl1: tx underrun, increasing tx start threshold to 420 bytes

Finally I apologize if this is posted in the wrong forum, I'm not sure where this should have gone

Hi, i have the same problem (sortof) but with pfsense...
this occurs on heavy traffic / bittorrent abuse...
from what i've been reading, maybe it's the NIC

You can set mono to use an external syslog server, then you'll have your logs after the lockup. Google shows ample howtos on setting up a syslog server with your OS of choice.

You may also want to look at your RAM usage, as 64 MB is the minimum required.

If you have a lot of bittorrent and the like running, then you could be overflowing your connection state table too. Check the status.php page for a look at the state table. Mono's limit is 30000.

db

Just FYI- Under Freebsd those 'xl'  3c0m nics will "fine tune" themselves when they see traffic increase. Its actually normal to see them do it after a boot and a heavy load from traffic. It should not keep happening over and over or show other errors. We have several machines running webservers or as mailservers and after a boot they will slowly crank up with these "increasing tx start threshold " as needed.