network slows 1000x
Hello, I have a Duron based FreeBSD box (4.10-current) that I am using as a nat box. Kernel is recompiled with options. When a lot of traffic flows across the interface from internal to external networks, I find that the natd appears to get congested, making ping times to the next hop (still on my network) go from .1 ms to 1000s of ms. Rebooting fixes it Nothing is reported in dmesg (I upgraded from 4.6 where I saw this problem) then there would be this message about queue lengths... The relavent devices are... xl0: 3Com 3c905-TX Fast Etherlink XL port 0xc800-0xc83f irq 5 at device 11.0 on pci0 vr0: VIA VT6102 Rhine II 10/100BaseTX port 0xe400-0xe4ff mem 0xe0117000-0xe01170ff irq 11 at device 18.0 on pci0 This happens once a day there abouts (depends on what's been moved between internal and external networks. Any ideas? Blake Freeburg ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: network slows 1000x
Blake Freeburg wrote: When a lot of traffic flows across the interface from internal to external networks, I find that the natd appears to get congested, making ping times to the next hop (still on my network) go from .1 ms to 1000s of ms. Rebooting fixes it --- cut - This happens once a day there abouts (depends on what's been moved between internal and external networks. i've seen a box with 2 NICs where 1 NIC needed a reboot every 2 or 3 days (replaced it when the problem was clear), i think the logfiles (on M0n0wall) mentioned corrupted memory on the NIC, so perhaps it's due to failure of one of your networkcards ? ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
