Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 18 Aug 2013 20:22:49 +0200
From:      Willem Jan Withagen <wjw@digiware.nl>
To:        net@freebsd.org
Subject:   Getting resets for em0 device
Message-ID:  <521110F9.1000906@digiware.nl>

next in thread | raw e-mail | index | archive | help
Hi,

I'm sending a zpool from one server to in neighbor.

Send size:
zfs send -v -R  tank@now \
        | mbuffer -4 -m 100M -O box:6666

Receive side:
mbuffer -4 -m 1000M -I 6666 \
        | zfs receive -e -v tank

This keep the 1Gbit link between them reasonably full. Most of the time
mbuffer reports 110 MiB/sec....

But on the receiver side I'm getting aborts. Log file is below.

Now the first question is:
	Does the switch reset?
		It's an HP 1800-24G
	Does FreeBSD reset?
		Because of the watchdog timeout?
		Because of too high packet rates?

And if it is FreeBSD, what can be done to fix it.

Version:
FreeBSD box.digiware.nl 9.2-PRERELEASE FreeBSD 9.2-PRERELEASE #3
r254428: Fri Aug 16 23:50:28 CEST 2013

Motherboard:
System Information
        Manufacturer: Supermicro
        Product Name: C2SBX

Ehternet controller:
em0@pci0:0:25:0:        class=0x020000 card=0x10bd15d9 chip=0x10bd8086
rev=0x02 hdr=0x00
    vendor     = 'Intel Corporation'
    device     = '82566DM-2 Gigabit Network Connection'
    class      = network
    subclass   = ethernet



dmesg, dmidecode, kernelconfig, full pciconf on:
	http://www.tegenbosch28.nl/FreeBSD/Systems/BOX

Thanx for any suggestions,
--WjW

Aug 18 18:24:55 box kernel: em0: Watchdog timeout -- resetting
Aug 18 18:24:55 box kernel: em0: Queue(0) tdh = 1688, hw tdt = 1743
Aug 18 18:24:55 box kernel: em0: TX(0) desc avail = 4031,Next TX to
Clean = 1678
Aug 18 18:24:55 box kernel: em0: Link is Down
Aug 18 18:24:55 box kernel: em0: link state changed to DOWN
Aug 18 18:24:57 box kernel: em0: Link is up 1000 Mbps Full Duplex
Aug 18 18:24:57 box kernel: em0: link state changed to UP
Aug 18 18:32:15 box kernel: em0: Watchdog timeout -- resetting
Aug 18 18:32:15 box kernel: em0: Queue(0) tdh = 2402, hw tdt = 2453
Aug 18 18:32:15 box kernel: em0: TX(0) desc avail = 4045,Next TX to
Clean = 2402
Aug 18 18:32:15 box kernel: em0: Link is Down
Aug 18 18:32:15 box kernel: em0: link state changed to DOWN
Aug 18 18:32:17 box kernel: em0: Link is up 1000 Mbps Full Duplex
Aug 18 18:32:17 box kernel: em0: link state changed to UP
Aug 18 18:33:00 box kernel: Limiting closed port RST response from 856
to 200 packets/sec
Aug 18 19:23:29 box kernel: em0: Watchdog timeout -- resetting
Aug 18 19:23:29 box kernel: em0: Queue(0) tdh = 613, hw tdt = 616
Aug 18 19:23:29 box kernel: em0: TX(0) desc avail = 4093,Next TX to
Clean = 613
Aug 18 19:23:29 box kernel: em0: Link is Down
Aug 18 19:23:29 box kernel: em0: link state changed to DOWN
Aug 18 19:23:31 box kernel: em0: Link is up 1000 Mbps Full Duplex
Aug 18 19:23:31 box kernel: em0: link state changed to UP



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?521110F9.1000906>