Date: Wed, 30 May 2001 10:43:31 +1000 From: Ralph Seberry <mischief@lanesbry.com> To: Peter <fbsdq@yahoo.com> Cc: freebsd-questions@freebsd.org Subject: Re: win2k Networking very slow Message-ID: <20010530104331.B7828@lanesbry.com> In-Reply-To: <SAK.2001.05.29.enecqpnb@support10>; from fbsdq@yahoo.com on Tue, May 29, 2001 at 08:31:23AM -0600 References: <SAK.2001.05.29.enecqpnb@support10>
next in thread | previous in thread | raw e-mail | index | archive | help
--AhhlLboLdkugWU4S
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
I wrote to Peter about his question of a slow network connection
to win95, because I have a similar problem. I think I have peter's
solution but I'm stuck.
My position:
Box A: win2k + intel etherexpress Pro 100B NIC, speed to internet is
much better than connectivity to box B via samba/ftp.
Box B: freebsd 4.1 + netgear 100 baseTx, speed to internet fine.
I've looked at several avenues (see below), but I'm missing something.
The connection just "pauses" frequently. (no lights on switch flicker.)
On Tuesday, 29 May 2001 at 08:31, Peter <fbsdq@yahoo.com> wrote:
> > Did you get a response to this, I have a similar problem
> No response, thinking about buying new NIC's.
I tried that :-(
I've done some more research now. See, I had win98 and it would
slow down intermittently. But I had to upgrade to win2k (for work)
and it is 3 orders of magnitude too slow and hangs a lot! (not
collisions actually, from looking at the switch.)
Someone else on freebsd-questions got advised to turn of delayed ack
(which is funny, because I think delayed ack in win2k is part of my
problem) by:
sysctl -w net.inet.ip.delayed_ack=0
which gave an order of magnitude performance increase.
Someone else said collisions could be half duplex NIC,
so I checked. That's not it. In fact I might have that from
the samba list. They said if one side is 1/2 duplex, one full,
then one side gets lots of stalls, the other lots of dropped
packets, iirc.
I found a microsoft knowledgebase article on this:
Q169789 which complains the interframe gap is smaller than
the 802.3 specification (so it's not a microsoft bug, natch)
but gives a work-around. Sounds like Peter's situtation.
As for my problem, after turning off delayed_ack, I'm still
2 orders of magnitude slower than expected (up from 600bytes/sec
to 6k/second). But at least it doesn't hang any more.
This problem exists on ftp as well as samba.
I turned of delayed ack in win2k (called SackOpts - special ack opts -
in microsoft knowledgebase article Q120642 and Q224829 on microsoft
tcpip tuning) and I think it made some difference, but it's subjective.
I found knowledgebase article Q244826 on "slow msdos file copy",
which suggests changing TCPWindow, but that didn't work. The
description (of losing packets due to incompatible packing of the
TCP window) sounds plausible for my symptoms, though.
I tried fiddling with rfc1323, ttl, recvspace and sendspace parameters
to little avail. (These were suggested in a Nov99 post to the samba
list, although there it was a setsockopts() mod to samba code.)
sysctl -w net.inet.ip.ttl=128
sysctl -w net.inet.ip.rfc1323=0
sysctl -w net.inet.ip.sendspace=32768
sysctl -w net.inet.ip.recvspace=32768
I tweaked the same paramaters in win2k. to no avail.
suggestions from the list welcome. I'm appending ifconfig,
dmesg to entice you, but sysctl -A was 15k, so ask for
what you want out of that.
--AhhlLboLdkugWU4S
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="ifconfig.out"
dc0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
inet X.X.X.10 netmask 0xffffff00 broadcast X.X.X.255
inet6 fe80::2a0:ccff:fed0:2c5d%dc0 prefixlen 64 scopeid 0x1
ether 00:a0:cc:d0:2c:5d
media: autoselect (100baseTX <full-duplex>) status: active
supported media: autoselect 100baseTX <full-duplex> 100baseTX 10baseT/UTP <full-duplex> 10baseT/UTP none
--AhhlLboLdkugWU4S
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="dmesg.out"
Copyright (c) 1992-2000 The FreeBSD Project.
Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
The Regents of the University of California. All rights reserved.
FreeBSD 4.1-RELEASE #3: Mon Aug 28 21:02:42 EST 2000
root@kryten:/usr/src/sys/compile/KRYTEN
Timecounter "i8254" frequency 1193182 Hz
CPU: Pentium/P54C (200.46-MHz 586-class CPU)
Origin = "GenuineIntel" Id = 0x52c Stepping = 12
Features=0x1bf<FPU,VME,DE,PSE,TSC,MSR,MCE,CX8>
real memory = 268435456 (262144K bytes)
avail memory = 257679360 (251640K bytes)
Preloaded elf kernel "kernel" at 0xc0361000.
Preloaded userconfig_script "/boot/kernel.conf" at 0xc036109c.
Intel Pentium detected, installing workaround for F00F bug
md0: Malloc disk
npx0: <math processor> on motherboard
npx0: INT 16 interface
pcib0: <Host to PCI bridge> on motherboard
pci0: <PCI bus> on pcib0
isab0: <Intel 82371AB PCI to ISA bridge> at device 7.0 on pci0
isa0: <ISA bus> on isab0
atapci0: <Intel PIIX4 ATA33 controller> port 0xf000-0xf00f at device 7.1 on pci0
ata0: at 0x1f0 irq 14 on atapci0
ata1: at 0x170 irq 15 on atapci0
pci0: <Intel 82371AB/EB (PIIX4) USB controller> at 7.2 irq 11
chip1: <Intel 82371AB Power management controller> port 0x6200-0x620f at device 7.3 on pci0
pci0: <Matrox MGA Millennium 2064W graphics accelerator> at 18.0 irq 10
dc0: <82c169 PNIC 10/100BaseTX> port 0x6400-0x64ff mem 0xe0804000-0xe08040ff irq 9 at device 20.0 on pci0
dc0: Ethernet address: 00:a0:cc:d0:2c:5d
miibus0: <MII bus> on dc0
ukphy0: <Generic IEEE 802.3u media interface> on miibus0
ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
fdc0: <NEC 72065B or clone> at port 0x3f0-0x3f5,0x3f7 irq 6 drq 2 on isa0
fdc0: FIFO enabled, 8 bytes threshold
fd0: <1440-KB 3.5" drive> on fdc0 drive 0
atkbdc0: <Keyboard controller (i8042)> at port 0x60,0x64 on isa0
atkbd0: <AT Keyboard> flags 0x1 irq 1 on atkbdc0
kbd0 at atkbd0
vga0: <Generic ISA VGA> at port 0x3c0-0x3df iomem 0xa0000-0xbffff on isa0
sc0: <System console> at flags 0x100 on isa0
sc0: VGA <16 virtual consoles, flags=0x300>
sio0 at port 0x3f8-0x3ff irq 4 flags 0x10 on isa0
sio0: type 16550A
sio1 at port 0x2f8-0x2ff irq 3 on isa0
sio1: type 16550A
ppc0: <Parallel port> at port 0x378-0x37f irq 7 on isa0
ppc0: Generic chipset (EPP/NIBBLE) in COMPATIBLE mode
ppi0: <Parallel I/O> on ppbus0
lpt0: <Printer> on ppbus0
lpt0: Interrupt-driven port
plip0: <PLIP network interface> on ppbus0
ad0: 19609MB <QUANTUM FIREBALLP KX20.5> [39842/16/63] at ata0-master using UDMA33
ad2: 28629MB <QUANTUM FIREBALLlct15 30> [58168/16/63] at ata1-master using UDMA33
Mounting root from ufs:/dev/ad0s1a
uhci0: <Intel 82371AB/EB (PIIX4) USB controller> port 0x6300-0x631f irq 11 at device 7.2 on pci0
usb0: <Intel 82371AB/EB (PIIX4) USB controller> on uhci0
usb0: USB revision 1.0
uhub0: Intel UHCI root hub, class 9/0, rev 1.00/1.00, addr 1
uhub0: 2 ports with 2 removable, self powered
uhub0: port 1 power on failed, IOERROR
uhub0: port 2 power on failed, IOERROR
--AhhlLboLdkugWU4S--
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010530104331.B7828>