I have a problem whose solutions I have yet to find and would greatly appreciate it if someone out there knew and could share the answer with me. My network card is a 3com 3c905c (I think). I have installed Red Hat Linux 7.2 on an HP Vectra 600mhz for the purposes of running a game server (Return to Castle Wolfenstein if you want to know). The problem is that after I boot the machine I can ping, ssh, ftp, etc to the server from any machine until about 5-15 minutes of inactivity have passed, whereupon I find that I can no longer reach the server. It hasn't crashed. I can get a local tty (console) and from there I can ping out to my default gateway, after which it is reachable for another 5-15 minutes. I am not running ipchains or any other firewall and I have disabled the apmd (advanced power management daemon) in an effort to troubleshoot this bizarre problem. I have another machine of the same hardware model running Mandrake 8.0 without this problem. I have also written and run a simple script on the new Red Hat 7.2 box that loops perpetually, pinging another computer once every minute, but now from home I find that I can no longer reach it. I expect that I will be able to go in tomorrow and find that it has not crashed, just stopped responding to network requests (not even ARPs). Any help would be appreciated. Do I have a flaky driver or a bad NIC?

your nic is screwed. i searched google, apparantly many other people have the same problem. choice #1 try to find a patch choice #2 buy a new nic found this article: http://www.tux.org/hypermail/linux-vortex/2001-Oct/0017.html is it using rtl8139.o? this is the same driver i am using with a non-3com card. i believe it should use a 3c* driver instead of an rt* driver. i have a 3c*9* card which uses the 3c* module. (the * means i am too lazy to type out the full model name) // timothy

I'm the president of the University of Akron Amateur Radio club. We are currently trying to get a new web server together with RedHat 7.2, and have had the same exact problem. A club member had the same problem on completely different hardware. We have even tried new network drivers, but that doesn't seem to work. Going to try and recompile the kernel tonight, and update RedHat. Please feel free to mail any suggestions to me - this problem really has us all stumped.

I found a fix - at least for our machine, and it was all related to APM. I followed you and killed the apmd, and removed it from the startup options. That didn't seem to work, but I then completely disabled APM int he BIOS, and the combination of those two things has resulted in a machine that has been up for over a day, compared to the 15 minutes it used to see. Best of luck!