When I use ipchains to MASQ there seems to be a "Buffer" that fills up after a day or so that causes my external nic to lose connection with my ISP (I have DSL and a Cisco 678 router). I'm running 2 Red Hat supported NIC's. Internet side nic is a 10 and the Network side is a 100. I'm running the latest SMP kernel from RHN. After about a day or so of GREAT performance, the external NIC seems to "Bog" down. I still have local network and it runs fine,for a while, then it too eventually dies. This does not happen when I only run with one NIC.
Example... I will be able to ping another host on the Internet and get a 105 ms reply from the server on the Internet connected NIC. Then, after some usage from the local network, or a good day's wait, the ping from the Internet host will go up to 200, then 600, then into the thousands. Eventually, I actually lose WAN connection. BUT, I have local network for a short time after this happens, then, I lose that as well and the server, if left unattended, will lock up and require a reboot. What I have to do is /etc/init.d/network restart then it to go back to normal as long as I can catch it before the server locks up. I guess I can create a cron job to do that nightly (The network restart)... Any one know what could cause this. I have tired several NIC's and even reloaded my system, and this problem ALWAYS comes back... I am using the Current SMP kernel from RHN as of this last weekend (As well as all the current patches up 2 date). This happened on the prior kernel as well. If you need to know any additional software that I am using, let me know... Is is a quick rundown of what I'm using... SAMBA APACHE DHCPD ProFTP SNORT NAGIOS Any Help would be appreciated ... Thanks Joe -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED]?subject=unsubscribe https://listman.redhat.com/mailman/listinfo/redhat-list
