Re: Writes to samba server very, very slow (SOLVED)
Try disabling apm.
# config -e -o /nbsd /bsd
ukc> disable apm
252 apm0 disabled
ukc> quit
# cp /bsd /obsd
# mv /nbsd /bsd
# reboot
If that speeds it up you have the "hlt hlt" issue. it's fixed in
current and stable
-Bob
* Gary Clemans-Gibbon <[EMAIL PROTECTED]> [2005-07-22 18:14]:
> Gary Clemans-Gibbon wrote:
> >Hi All,
> >
> >I just built a OpenBSD 3.7 samba file server for my home lan. It's a P3
> >500, 128mb RAM, with a 2 gig IDE HDD for the OS and two x Maxtor 200 GB
> >IDE drives for data.
> >
> >Everything is working fine except that when I copy files to the box from
> >a Windows XP box the transfers are very slow, like 9 minutes for a 48 Mb
> >file. Copying the same file back to the win box is quick - a couple of
> >seconds as you'd expect.
> >
> >Please forgive me if I don't provide all the needed info here or if I
> >didn't run any obvious checks. Please indicate what info is needed and
> >how to get it and I'll repost it.
> >
> >This same hardware was previously running RH7.3 with samba and worked
> >fine. I've tried a different ethernet cable and a different port on my
> >switch too.
> >
> >many thanks in advance,
> >Gary
> >
> >Here is /etc/samba/smb.conf global section..
> >
> >[global]
> >workgroup = myworkgroup
> >server string = My Samba Server
> >hosts allow = 192.168.20. 127.0.0.1
> >log file = /var/log/smbd.%m
> >security = user
> >socket options = TCP_NODELAY IPTOS_LOWDELAY
> >read raw = yes
> >write raw = yes
> >
> >Here is dmesg
> >
> >OpenBSD 3.7 (GENERIC) #50: Sun Mar 20 00:01:57 MST 2005
> >[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
> >cpu0: Intel Pentium III ("GenuineIntel" 686-class, 512KB L2 cache) 501 MHz
> >cpu0:
> >FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER,MMX,FXSR,SSE
> >
> >
> >cpu0: disabling processor serial number
> >real mem = 133787648 (130652K)
> >avail mem = 115580928 (112872K)
> >using 1658 buffers containing 6791168 bytes (6632K) of memory
> >mainbus0 (root)
> >bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @ 0xf0520
> >apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
> >apm0: APM power management enable: unrecognized device ID (9)
> >apm0: APM engage (device 1): power management disabled (1)
> >apm0: AC on, battery charge unknown
> >pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
> >pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
> >pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev 0x00)
> >pcibios0: PCI bus #1 is the last bus
> >bios0: ROM list: 0xc/0x8000
> >cpu0 at mainbus0
> >pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
> >pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
> >ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
> >cpu0:
> >FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER,MMX,FXSR,SSE
> >
> >
> >cpu0: disabling processor serial number
> >real mem = 133787648 (130652K)
> >avail mem = 115580928 (112872K)
> >using 1658 buffers containing 6791168 bytes (6632K) of memory
> >mainbus0 (root)
> >bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @ 0xf0520
> >apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
> >apm0: APM power management enable: unrecognized device ID (9)
> >apm0: APM engage (device 1): power management disabled (1)
> >apm0: AC on, battery charge unknown
> >pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
> >pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
> >pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev 0x00)
> >pcibios0: PCI bus #1 is the last bus
> >bios0: ROM list: 0xc/0x8000
> >cpu0 at mainbus0
> >pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
> >pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
> >ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
> >pci1 at ppb0 bus 1
> >vga1 at pci1 dev 0 function 0 "Nvidia Riva TNT" rev 0x04
> >wsdisplay0 at vga1: console (80x25, vt100 emulation)
> >wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
> >pcib0 at pci0 dev 4 function 0 "Intel 82371AB PIIX4 ISA" rev 0x02
> >pciide0 at pci0 dev 4 function 1 "Intel 82371AB IDE" rev 0x01: DMA,
> >channel 0 wired to compatibility, channel 1 wired to compatibility
> >wd0 at pciide0 channel 0 drive 0:
> >wd0: 16-sector PIO, LBA, 1032MB, 2114180 sectors
> >wd1 at pciide0 channel 0 drive 1:
> >wd1: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
> >wd0(pciide0:0:0): using PIO mode 4, DMA mode 2
> >wd1(pciide0:0:1): using PIO mode 4, Ultra-DMA mode 2
> >wd2 at pciide0 channel 1 drive 0:
> >wd2: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
> >atapiscsi0 at pciide0 channel 1 drive 1
> >scsibus0 at atapiscsi0: 2 targets
> >cd0 at scsibus0 targ 0 lun 0: <, 52X24X52 CD-RW, 1.07> SCSI0 5/cdrom
> >removable
> >wd2(pciide0:1:0): using PIO
Re: Writes to samba server very, very slow (SOLVED)
Gary Clemans-Gibbon wrote:
Hi All,
I just built a OpenBSD 3.7 samba file server for my home lan. It's a P3
500, 128mb RAM, with a 2 gig IDE HDD for the OS and two x Maxtor 200 GB
IDE drives for data.
Everything is working fine except that when I copy files to the box from
a Windows XP box the transfers are very slow, like 9 minutes for a 48 Mb
file. Copying the same file back to the win box is quick - a couple of
seconds as you'd expect.
Please forgive me if I don't provide all the needed info here or if I
didn't run any obvious checks. Please indicate what info is needed and
how to get it and I'll repost it.
This same hardware was previously running RH7.3 with samba and worked
fine. I've tried a different ethernet cable and a different port on my
switch too.
many thanks in advance,
Gary
Here is /etc/samba/smb.conf global section..
[global]
workgroup = myworkgroup
server string = My Samba Server
hosts allow = 192.168.20. 127.0.0.1
log file = /var/log/smbd.%m
security = user
socket options = TCP_NODELAY IPTOS_LOWDELAY
read raw = yes
write raw = yes
Here is dmesg
OpenBSD 3.7 (GENERIC) #50: Sun Mar 20 00:01:57 MST 2005
[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel Pentium III ("GenuineIntel" 686-class, 512KB L2 cache) 501 MHz
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER,MMX,FXSR,SSE
cpu0: disabling processor serial number
real mem = 133787648 (130652K)
avail mem = 115580928 (112872K)
using 1658 buffers containing 6791168 bytes (6632K) of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @ 0xf0520
apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: APM engage (device 1): power management disabled (1)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev 0x00)
pcibios0: PCI bus #1 is the last bus
bios0: ROM list: 0xc/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER,MMX,FXSR,SSE
cpu0: disabling processor serial number
real mem = 133787648 (130652K)
avail mem = 115580928 (112872K)
using 1658 buffers containing 6791168 bytes (6632K) of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @ 0xf0520
apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: APM engage (device 1): power management disabled (1)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev 0x00)
pcibios0: PCI bus #1 is the last bus
bios0: ROM list: 0xc/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
pci1 at ppb0 bus 1
vga1 at pci1 dev 0 function 0 "Nvidia Riva TNT" rev 0x04
wsdisplay0 at vga1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
pcib0 at pci0 dev 4 function 0 "Intel 82371AB PIIX4 ISA" rev 0x02
pciide0 at pci0 dev 4 function 1 "Intel 82371AB IDE" rev 0x01: DMA,
channel 0 wired to compatibility, channel 1 wired to compatibility
wd0 at pciide0 channel 0 drive 0:
wd0: 16-sector PIO, LBA, 1032MB, 2114180 sectors
wd1 at pciide0 channel 0 drive 1:
wd1: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
wd0(pciide0:0:0): using PIO mode 4, DMA mode 2
wd1(pciide0:0:1): using PIO mode 4, Ultra-DMA mode 2
wd2 at pciide0 channel 1 drive 0:
wd2: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
atapiscsi0 at pciide0 channel 1 drive 1
scsibus0 at atapiscsi0: 2 targets
cd0 at scsibus0 targ 0 lun 0: <, 52X24X52 CD-RW, 1.07> SCSI0 5/cdrom
removable
wd2(pciide0:1:0): using PIO mode 4, Ultra-DMA mode 2
cd0(pciide0:1:1): using PIO mode 4, Ultra-DMA mode 2
uhci0 at pci0 dev 4 function 2 "Intel 82371AB USB" rev 0x01: irq 9
usb0 at uhci0: USB revision 1.0
uhub0 at usb0
uhub0: Intel UHCI root hub, class 9/0, rev 1.00/1.00, addr 1
uhub0: 2 ports with 2 removable, self powered
"Intel 82371AB Power Mgmt" rev 0x02 at pci0 dev 4 function 3 not configured
dc0 at pci0 dev 9 function 0 "ADMtek AN983" rev 0x11: irq 9, address
00:50:bf:fe:83:bc
ukphy0 at dc0 phy 1: Generic IEEE 802.3u media interface
ukphy0: OUI 0x000749, model 0x0001, rev. 1
isa0 at pcib0
isadma0 at isa0
pckbc0 at isa0 port 0x60/5
pckbd0 at pckbc0 (kbd slot)
pckbc0: using irq 1 for kbd s
Re: Writes to samba server very, very slow
On 7/19/05, Gary Clemans-Gibbon <[EMAIL PROTECTED]> wrote: > Forgot about the /dev/null idea. interesting result. I scp'd a 10 Mb > file from my gentoo box and it completed fast in a few seconds - speed > 3.3 Mb/s. Not great but faster than the other experiences. > > I then did the same with a 2.5Gb file and it started off at 3.3 or 3.5 > Mb/sec but tailed of continuously until by the time the transfer was 12% > done the transfer speed was down to 50 kb/sec. heres a clue: have done anything with Nick's suggestions yet?
Re: Writes to samba server very, very slow
On Tue, 19 Jul 2005, Gary Clemans-Gibbon wrote: > Darren Tucker wrote: >> Gary Clemans-Gibbon wrote: >> >>> Also I have just grabbed the stable branch from cvs and am running stable >>> GENERIC and still doesn't fix it. Just a recap - the problem is not just >>> samba writes to either of the data disks from the network via samba or scp >>> are painfully slow. Reads from the box to the network are fast. Copying >>> files from one disk to another one the server are fast. >> How does Samba's logs look like, smbd.smbd etc.? Also, look if there's any extra smbd processes running. -- Hannu Pysdys | [EMAIL PROTECTED]
Re: Writes to samba server very, very slow
Hi, ...on Tue, Jul 19, 2005 at 12:35:37PM -0400, Michael Shalayeff wrote: > Making, drinking tea and reading an opus magnum from Alexander Bochmann: > > Just to drop this in again - at least here, writing from > > an OS X (10.2 or 10.3) system to a Samba 2 server is > > abysmally slow up to the point that OS X claims to have > > lost connection to the server, regardless of whether > > the Samba server runs on OpenBSD or Linux. > it sounds as you have duplex mismatch on one of the ifaces... Yes, I suspected that too, but other applications, like rsync for example, run at normal speed between the affected systems. > > yet, though... Also didn't really try Samba 3 yet, as > > it didn't play nice with my OS/2 client last time... > on the contrary they have fixed some memory leaks and it > does not crash anymore in daemon mode as it used to do... Don't know, but the OS/2 client had trouble with it's extended attributes on the Samba server - which Samba 3 claims to support, but it didn't work for me. But that's a bit offtopic :) Alex.
Re: Writes to samba server very, very slow
Darren Tucker wrote: Gary Clemans-Gibbon wrote: Also I have just grabbed the stable branch from cvs and am running stable GENERIC and still doesn't fix it. Just a recap - the problem is not just samba writes to either of the data disks from the network via samba or scp are painfully slow. Reads from the box to the network are fast. Copying files from one disk to another one the server are fast. Taking the disks out of the equation entirely: does scp'ing a file to /dev/null run fast or slow? $ dmesg [...] uhci0 at pci0 dev 4 function 2 "Intel 82371AB USB" rev 0x01: irq 9 [...] dc0 at pci0 dev 9 function 0 "ADMtek AN983" rev 0x11: irq 9, address The NIC is sharing IRQ 9 with the USB controller, can you select another IRQ for the NIC in the BIOS? Especially if you can select a non-shared one? Darren, Forgot about the /dev/null idea. interesting result. I scp'd a 10 Mb file from my gentoo box and it completed fast in a few seconds - speed 3.3 Mb/s. Not great but faster than the other experiences. I then did the same with a 2.5Gb file and it started off at 3.3 or 3.5 Mb/sec but tailed of continuously until by the time the transfer was 12% done the transfer speed was down to 50 kb/sec. Very strange. Something is not happy here. Does this info give you any clues? I'm more baffled than before! thanks Gary
Re: Writes to samba server very, very slow
Darren Tucker wrote: Gary Clemans-Gibbon wrote: Also I have just grabbed the stable branch from cvs and am running stable GENERIC and still doesn't fix it. Just a recap - the problem is not just samba writes to either of the data disks from the network via samba or scp are painfully slow. Reads from the box to the network are fast. Copying files from one disk to another one the server are fast. Taking the disks out of the equation entirely: does scp'ing a file to /dev/null run fast or slow? $ dmesg [...] uhci0 at pci0 dev 4 function 2 "Intel 82371AB USB" rev 0x01: irq 9 [...] dc0 at pci0 dev 9 function 0 "ADMtek AN983" rev 0x11: irq 9, address The NIC is sharing IRQ 9 with the USB controller, can you select another IRQ for the NIC in the BIOS? Especially if you can select a non-shared one? Good suggestion Darren, I thought you might have hit upon it there. I remember a problem from way back when a promise raid controller card was fussy about which pci slot it was in. Anyway I tried moving the NIC to another slot and also manually set the irq in the bios to 12. Neither made any difference though # dmesg | grep irq uhci0 at pci0 dev 4 function 2 "Intel 82371AB USB" rev 0x01: irq 9 dc0 at pci0 dev 10 function 0 "ADMtek AN983" rev 0x11: irq 12, address 00:50:bf:9c:62:e4 pckbc0: using irq 1 for kbd slot lpt0 at isa0 port 0x378/4 irq 7 pccom0 at isa0 port 0x3f8/8 irq 4: ns16550a, 16 byte fifo pccom1 at isa0 port 0x2f8/8 irq 3: ns16550a, 16 byte fifo fdc0 at isa0 port 0x3f0/6 irq 6 drq 2 sb1 at isapnp0 "Creative SB AWE64 PnP, CTL0042, , Audio" port 0x220/16,0x330/2,0x388/4 irq 5 drq 1,5: dsp v4.16 wdc2 at isapnp0 "Creative SB AWE64 PnP, CTL2011, PNP0600, IDE" port 0x168/8,0x36e/2 irq 10
Re: Writes to samba server very, very slow
Gary Clemans-Gibbon wrote: Also I have just grabbed the stable branch from cvs and am running stable GENERIC and still doesn't fix it. Just a recap - the problem is not just samba writes to either of the data disks from the network via samba or scp are painfully slow. Reads from the box to the network are fast. Copying files from one disk to another one the server are fast. Taking the disks out of the equation entirely: does scp'ing a file to /dev/null run fast or slow? $ dmesg [...] uhci0 at pci0 dev 4 function 2 "Intel 82371AB USB" rev 0x01: irq 9 [...] dc0 at pci0 dev 9 function 0 "ADMtek AN983" rev 0x11: irq 9, address The NIC is sharing IRQ 9 with the USB controller, can you select another IRQ for the NIC in the BIOS? Especially if you can select a non-shared one? -- Darren Tucker (dtucker at zip.com.au) GPG key 8FF4FA69 / D9A3 86E9 7EEE AF4B B2D4 37C9 C982 80C7 8FF4 FA69 Good judgement comes with experience. Unfortunately, the experience usually comes from bad judgement.
Re: Writes to samba server very, very slow
John Brooks wrote: How about a nic from a different mfr? Using another good 'dc' nic doesn't rule out a basic hardware incompatibility related directly to that brand of nic card coupled with your other hardware. JB . Even though it worked fine with RH7.3 a three + year old OS ? I've seen specific nic cards conflict with switches, hubs, and routers. Mix in some driver code that may do things a little different from one OS to another... By your response I take it you did swap out with the same brand of nic. Might be worth a try, what else has worked? . I've tried two different makes of NIC as can be seen from the two different dmesg listings I've posted in two seperate messages... The first was an SMC 1244TX and the second (and currently installed is a NIC that I have bought in the past in bulk and have proven very reliable with bsd, linux and windows, Encore ENL832-TX+ based on the RTL8139 chip. This NIC as far as I know is as generic as anything and I've used dozens of them in the past in all kinds of systems without a hitch. It is also listed on the Hardware compatability list. http://www.openbsd.org/i386.html#hardware Gary
Re: Writes to samba server very, very slow - more info
Gary Clemans-Gibbon wrote: > I've tried a couple of things. Firstly I swapped out the NIC for a > different brand, no change. Writes god-awful slow, reads nice and zippy. When you swapped out the NIC, did you try a non-dc NIC? Maybe a different slot? > $ netstat -in > dc0 1500 192.168.20/ 192.168.20.3 24956 040617 0 0 In general, I've found that mismatched duplex will cause excessive collisions. You don't seem to have that here. > $ netstat -ss > ip: > 24911 total packets received > 16 bad header checksums > 186 with data size < data length > 24709 packets for this host > 40602 packets sent from this host The bad checksums seem odd. I checked all my machines here, and none had IP header checksum errors, even facing the Internet. Is there any chance you could borrow a pair of good gig network cards to test with? Belkin sells some cheap sk cards, they might be a good choice if you want to purchase them. At least try changing the NIC in your XP machine if you haven't already. Keep doing the tests with scp, that keeps Samba out of your diagnosis.
Re: Writes to samba server very, very slow
> > How about a nic from a different mfr? Using another good 'dc' nic doesn't > > rule out a basic hardware incompatibility related directly to that brand > > of nic card coupled with your other hardware. > > > > JB > > > > . > > > > Even though it worked fine with RH7.3 a three + year old OS ? I've seen specific nic cards conflict with switches, hubs, and routers. Mix in some driver code that may do things a little different from one OS to another... By your response I take it you did swap out with the same brand of nic. Might be worth a try, what else has worked?
Re: Writes to samba server very, very slow
> > Sounds like a bad nic. If you have a spare, you might try > swapping with it. > > > > Also, what's the other machine and what is it running? > > > > The NIC is fine, and yes I swapped it out early on as well as the cable > and the port on the switch. I've also tried a crossover cable. I've also > tried all the different media options in ifconfig as given by ifconfig > -m dc0. > > Gary > How about a nic from a different mfr? Using another good 'dc' nic doesn't rule out a basic hardware incompatibility related directly to that brand of nic card coupled with your other hardware. JB
Re: Writes to samba server very, very slow
On Tue, Jul 19, 2005 at 09:50:03PM -0400, Don Koch wrote: > [EMAIL PROTECTED] said: > > I heeded your words of wisdom and after some rummaging I found a crossover > > cable. Hooked it up the same winXP box direct and the exact same result. > > Sounds like a bad nic. If you have a spare, you might try swapping with it. > > Also, what's the other machine and what is it running? Ya, can you find say an fxp (Intel) and try that out instead?
Re: Writes to samba server very, very slow
Don Koch wrote: [EMAIL PROTECTED] said: I heeded your words of wisdom and after some rummaging I found a crossover cable. Hooked it up the same winXP box direct and the exact same result. Sounds like a bad nic. If you have a spare, you might try swapping with it. Also, what's the other machine and what is it running? The NIC is fine, and yes I swapped it out early on as well as the cable and the port on the switch. I've also tried a crossover cable. I've also tried all the different media options in ifconfig as given by ifconfig -m dc0. This is a basic install with only samba installed. Nothing could be simpler. Gary
Re: Writes to samba server very, very slow
[EMAIL PROTECTED] said: > I heeded your words of wisdom and after some rummaging I found a crossover > cable. Hooked it up the same winXP box direct and the exact same result. Sounds like a bad nic. If you have a spare, you might try swapping with it. Also, what's the other machine and what is it running? -- Don Koch http://www.krl.com/
Re: Writes to samba server very, very slow
Tim Hammerquist wrote: Gary Clemans-Gibbon wrote: Yes, I've tried different cables and different ports on the switch. This hardware has all been used before together. I cycled the power to the switch (can't find a reset button) and no change. Via samba or SCP it takes 7 minutes to write a file to the server and 12 seconds to read the same file. Tried an xover cable directly between the nodes? Tim . Tim, Just tried one. No difference. Also I am now running GENERIC stable. Still no change. Gary
Re: Writes to samba server very, very slow
Gary Clemans-Gibbon wrote: > Yes, I've tried different cables and different ports on the switch. > This hardware has all been used before together. I cycled the power to > the switch (can't find a reset button) and no change. Via samba or SCP > it takes 7 minutes to write a file to the server and 12 seconds to > read the same file. Tried an xover cable directly between the nodes? Tim
Re: Writes to samba server very, very slow
Avtar Gill wrote:
Stephen Marley wrote:
Have you tried a crossover cable, bypassing the switch?
Words of wisdom right there. I would definitely try a crossover cable
between the OpenBSD server and your client machine. That will show you
whether the problem lies with Samba or elsewhere. I remember
experiencing problems like yours years ago (no Samba though) and
specifying half-duplex resolved the issue.
Thanks Stephen.
I heeded your words of wisdom and after some rummaging I found a
crossover cable. Hooked it up the same winXP box direct and the exact
same result.
Also I have just grabbed the stable branch from cvs and am running
stable GENERIC and still doesn't fix it. Just a recap - the problem is
not just samba writes to either of the data disks from the network via
samba or scp are painfully slow. Reads from the box to the network are
fast. Copying files from one disk to another one the server are fast.
thanks
Gary
$ dmesg
OpenBSD 3.7-stable (GENERIC) #0: Tue Jul 19 17:45:55 PDT 2005
[EMAIL PROTECTED]:/home/krusadr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel Pentium III ("GenuineIntel" 686-class, 512KB L2 cache) 501 MHz
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER,MMX,FXSR,SSE
cpu0: disabling processor serial number
real mem = 133787648 (130652K)
avail mem = 115576832 (112868K)
using 1658 buffers containing 6791168 bytes (6632K) of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @ 0xf0520
apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: APM engage (device 1): power management disabled (1)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev 0x00)
pcibios0: PCI bus #1 is the last bus
bios0: ROM list: 0xc/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
pci1 at ppb0 bus 1
vga1 at pci1 dev 0 function 0 "Nvidia Riva TNT" rev 0x04
wsdisplay0 at vga1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
pcib0 at pci0 dev 4 function 0 "Intel 82371AB PIIX4 ISA" rev 0x02
pciide0 at pci0 dev 4 function 1 "Intel 82371AB IDE" rev 0x01: DMA,
channel 0 wired to compatibility, channel 1 wire
d to compatibility
wd0 at pciide0 channel 0 drive 0:
wd0: 16-sector PIO, LBA, 1032MB, 2114180 sectors
wd1 at pciide0 channel 0 drive 1:
wd1: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
wd0(pciide0:0:0): using PIO mode 4, DMA mode 2
wd1(pciide0:0:1): using PIO mode 4, Ultra-DMA mode 2
wd2 at pciide0 channel 1 drive 0:
wd2: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
atapiscsi0 at pciide0 channel 1 drive 1
scsibus0 at atapiscsi0: 2 targets
cd0 at scsibus0 targ 0 lun 0: <, 52X24X52 CD-RW, 1.07> SCSI0 5/cdrom
removable
wd2(pciide0:1:0): using PIO mode 4, Ultra-DMA mode 2
cd0(pciide0:1:1): using PIO mode 4, Ultra-DMA mode 2
uhci0 at pci0 dev 4 function 2 "Intel 82371AB USB" rev 0x01: irq 9
usb0 at uhci0: USB revision 1.0
uhub0 at usb0
uhub0: Intel UHCI root hub, class 9/0, rev 1.00/1.00, addr 1
uhub0: 2 ports with 2 removable, self powered
"Intel 82371AB Power Mgmt" rev 0x02 at pci0 dev 4 function 3 not configured
dc0 at pci0 dev 9 function 0 "ADMtek AN983" rev 0x11: irq 9, address
00:50:bf:9c:62:e4
ukphy0 at dc0 phy 1: Generic IEEE 802.3u media interface
ukphy0: OUI 0x000749, model 0x0001, rev. 1
isa0 at pcib0
isadma0 at isa0
pckbc0 at isa0 port 0x60/5
pckbd0 at pckbc0 (kbd slot)
pckbc0: using irq 1 for kbd slot
wskbd0 at pckbd0 (mux 1 ignored for console): console keyboard, using
wsdisplay0
pmsi0 at pckbc0 (aux slot)
pckbc0: using irq 12 for aux slot
wsmouse0 at pmsi0 mux 0
pcppi0 at isa0 port 0x61
midi0 at pcppi0:
sysbeep0 at pcppi0
lpt0 at isa0 port 0x378/4 irq 7
lm0 at isa0 port 0x290/8: W83781D
npx0 at isa0 port 0xf0/16: using exception 16
pccom0 at isa0 port 0x3f8/8 irq 4: ns16550a, 16 byte fifo
pccom1 at isa0 port 0x2f8/8 irq 3: ns16550a, 16 byte fifo
fdc0 at isa0 port 0x3f0/6 irq 6 drq 2
fd0 at fdc0 drive 0: 1.44MB 80 cyl, 2 head, 18 sec
isapnp0 at isa0 port 0x279: read port 0x203
sb1 at isapnp0 "Creative SB AWE64 PnP, CTL0042, , Audio" port
0x220/16,0x330/2,0x388/4 irq 5 drq 1,5: dsp v4.16
midi1 at sb1:
audio0 at sb1
opl0 at sb1: model OPL3
midi2 at opl0:
joy0 at isapnp0 "Creative SB AWE64 PnP, CTL7002, PNPB02F, Game" port 0x200/8
"Creative SB AWE64 PnP, CTL0022, , WaveTable" at isapnp0 port 0x620/4
not configured
wdc2 at isapnp0 "Creative SB AWE64 PnP, CTL2011, PNP0600, IDE" port
0x168/8,0x36e/2 irq 10
biomask ef45 netmask ef45 ttymask ffc7
pctr: 686-class user-level performance counters enabled
mtrr: Pentium Pro MTRR support
dkcsum: wd0 matched BIOS disk 80
dkcsum:
Re: Writes to samba server very, very slow
Stephen Marley wrote: Have you tried a crossover cable, bypassing the switch? Words of wisdom right there. I would definitely try a crossover cable between the OpenBSD server and your client machine. That will show you whether the problem lies with Samba or elsewhere. I remember experiencing problems like yours years ago (no Samba though) and specifying half-duplex resolved the issue. -- Avtar Gill | [EMAIL PROTECTED]
Re: Writes to samba server very, very slow
--On 19 July 2005 10:46 -0500, Daniel Ramaley wrote: I've had this problem before. Other protocols work OK (not great, but OK), but Samba is gets modem speeds on a fast ethernet connection. I've had poor speeds samba->windows before and sometimes been able to fix it by just disabling and re-enabling the network connection on the windows box (until the next time when it randomly happened again months later)...I suspect that, amongst other possible problems, there are some fairly broken nic drivers out there.
Re: Writes to samba server very, very slow
On Tue, Jul 19, 2005 at 02:57:33PM -0700, Gary Clemans-Gibbon wrote: > Yes, I've tried different cables and different ports on the switch. This > hardware has all been used before together. I cycled the power to the > switch (can't find a reset button) and no change. Via samba or SCP it > takes 7 minutes to write a file to the server and 12 seconds to read the > same file. Have you tried half-duplex at either 100 or 10Mb/s? Have you tried a different switch? Have you tried a crossover cable, bypassing the switch? Have you tried writing to a memory filesystem (to rule out HD)? -- stephen
Re: Writes to samba server very, very slow
I've had this problem before. Other protocols work OK (not great, but OK), but Samba is gets modem speeds on a fast ethernet connection. I've had some success in the past by manually specifying the network speed and duplexing. See "man hostname.if" for info on how to do that (pay particular attention to the example "dhcp media 100baseTX mediaopt full-duplex"). On Monday 18 July 2005 11:03 pm, Gary Clemans-Gibbon wrote: >Hi All, > >I just built a OpenBSD 3.7 samba file server for my home lan. It's a > P3 500, 128mb RAM, with a 2 gig IDE HDD for the OS and two x Maxtor > 200 GB IDE drives for data. > >Everything is working fine except that when I copy files to the box > from a Windows XP box the transfers are very slow, like 9 minutes for > a 48 Mb file. Copying the same file back to the win box is quick - a > couple of seconds as you'd expect. -- Dan Ramaley Digital Media Library Specialist (515) 271-1934 Cowles Library 140, Drake University
Re: Writes to samba server very, very slow
Stephen Marley wrote: On Tue, Jul 19, 2005 at 01:28:17PM -0700, Gary Clemans-Gibbon wrote: A little bit more info, i ran the following... . $ ifconfig -m dc0 dc0: flags=8843 mtu 1500 address: 00:50:bf:9c:62:e4 media: Ethernet autoselect (100baseTX full-duplex) status: active supported media: media none media 10baseT media 10baseT mediaopt full-duplex media 100baseTX media 100baseTX mediaopt full-duplex media autoselect inet 192.168.20.3 netmask 0xff00 broadcast 192.168.20.255 inet6 fe80::250:bfff:fe9c:62e4%dc0 prefixlen 64 scopeid 0x1 $ ifconfig dc0 media 100baseTX mediaopt full-duplex ifconfig: SIOCSIFMEDIA: Operation not permitted $ su Password: salyut# ksh # ifconfig dc0 media 100baseTX mediaopt full-duplex # ifconfig dc0 media 100baseTX # ifconfig dc0 media 10baseT # ifconfig dc0 media 100baseTX mediaopt full-duplex # ifconfig dc0 media 10baseT mediaopt full-duplex It speeded up for the 10baseT setting. I ran through the rest of the settings and now it seems hosed. I think I have thouroughly confused my switch now. Obviously it goes without saying that I want it to work at 100 not 10 but now it is really not working at all, like 30 minutes to copy a 5 Mb file in either direction! Not good :( Reset your switch. Try 100baseTX half-duplex. If autonegiotion is broken, and the switch is not manageable (to turn off autoneg), this may be the best you can do. Also, this may be obvious, but have you tried different cables? Yes, I've tried different cables and different ports on the switch. This hardware has all been used before together. I cycled the power to the switch (can't find a reset button) and no change. Via samba or SCP it takes 7 minutes to write a file to the server and 12 seconds to read the same file. thanks
Re: Writes to samba server very, very slow
On 7/19/05, Stephen Marley <[EMAIL PROTECTED]> wrote: > On Tue, Jul 19, 2005 at 01:28:17PM -0700, Gary Clemans-Gibbon wrote: > > A little bit more info, > > > > i ran the following... > > snip dont forget to use netstat -i (-e on windows) to look for errors on the line, which would be indicative of lan segment symptoms mentioned above. When you start a large file transfer which typically will run with the df bit set at the max mtu size you might see the error counters increment more. Also `sudo ping -D -i 0.05 -s 1472 x.x.x.x` works well for testing troublesome segments.
Re: Writes to samba server very, very slow
On Tue, Jul 19, 2005 at 01:28:17PM -0700, Gary Clemans-Gibbon wrote: > A little bit more info, > > i ran the following... > > > > . > > > $ ifconfig -m dc0 > > dc0: flags=8843 mtu 1500 > address: 00:50:bf:9c:62:e4 > media: Ethernet autoselect (100baseTX full-duplex) > status: active > supported media: > media none > media 10baseT > media 10baseT mediaopt full-duplex > media 100baseTX > media 100baseTX mediaopt full-duplex > media autoselect > inet 192.168.20.3 netmask 0xff00 broadcast 192.168.20.255 > inet6 fe80::250:bfff:fe9c:62e4%dc0 prefixlen 64 scopeid 0x1 > $ ifconfig dc0 media 100baseTX mediaopt full-duplex > ifconfig: SIOCSIFMEDIA: Operation not permitted > $ su > Password: > salyut# ksh > # ifconfig dc0 media 100baseTX mediaopt full-duplex > > # ifconfig dc0 media 100baseTX > # ifconfig dc0 media 10baseT > # ifconfig dc0 media 100baseTX mediaopt full-duplex > # ifconfig dc0 media 10baseT mediaopt full-duplex > > It speeded up for the 10baseT setting. I ran through the rest of the > settings and now it seems hosed. I think I have thouroughly confused my > switch now. Obviously it goes without saying that I want it to work at > 100 not 10 but now it is really not working at all, like 30 minutes to > copy a 5 Mb file in either direction! > > Not good :( Reset your switch. Try 100baseTX half-duplex. If autonegiotion is broken, and the switch is not manageable (to turn off autoneg), this may be the best you can do. Also, this may be obvious, but have you tried different cables? -- stephen
Re: Writes to samba server very, very slow
Garance A Drosihn wrote: At 11:15 PM -0700 7/18/05, Gary Clemans-Gibbon wrote: Thanks David, I just tried that line but it seems to be the same or if anything it seems even slower. I missed the start of this thread, but make sure that you do not have a duplex-mismatch with your ethernet cards. You want to be sure your server and your client(s) have the same duplex setting as whatever they are connected to (usually some switch). At one point we had excruciating performance problems with our samba server, and eventually this turned out to be a duplex mismatch between the server and the switch it was connected to. The switch was set to 100-Mb-full/duplex (which is to say, it was *not* doing autonegotiation), and the network interface on the server was set to autonegotiate. The server ended up with the wrong answer for duplex... Garance, Thanks for that. I am experiencing similar slow write speeds when I transfer uusing SCP as well so it's not samba. Copying a file from one HDD to the other on the samba box results in a fast write. It looks like communication. I've tried two NICs. The box is hooked to a 16 port switch which works fine. I don't know how to check the autonegotiation. $ ifconfig -a lo0: flags=8049 mtu 33224 inet 127.0.0.1 netmask 0xff00 inet6 ::1 prefixlen 128 inet6 fe80::1%lo0 prefixlen 64 scopeid 0x5 dc0: flags=8843 mtu 1500 address: 00:50:bf:9c:62:e4 media: Ethernet autoselect (100baseTX full-duplex) status: active inet 192.168.20.3 netmask 0xff00 broadcast 192.168.20.255 inet6 fe80::250:bfff:fe9c:62e4%dc0 prefixlen 64 scopeid 0x1 pflog0: flags=0<> mtu 33224 pfsync0: flags=0<> mtu 2020 enc0: flags=0<> mtu 1536 Is there someway I can try different manual settings with the NIC. I guess it is an ifconfig command. I'm off to google now. If you have any suggestion I'd be grateful. thanks Gary
Re: Writes to samba server very, very slow
Making, drinking tea and reading an opus magnum from Alexander Bochmann: > ...on Tue, Jul 19, 2005 at 12:56:03AM -0700, Tim Hammerquist wrote: > > > I've seen this same phenomenon when copying to from my OSX Powerbook and > > my fileserver (running both FreeBSD 5 and Gentoo Linux), with the OSX > > acting as samba client. > > Just to drop this in again - at least here, writing from > an OS X (10.2 or 10.3) system to a Samba 2 server is > abysmally slow up to the point that OS X claims to have > lost connection to the server, regardless of whether > the Samba server runs on OpenBSD or Linux. it sounds as you have duplex mismatch on one of the ifaces... > Couldn't be bothered to find out the reason for that > yet, though... Also didn't really try Samba 3 yet, as > it didn't play nice with my OS/2 client last time... on the contrary they have fixed some memory leaks and it does not crash anymore in daemon mode as it used to do... cu -- paranoic mickey (my employers have changed but, the name has remained)
Re: Writes to samba server very, very slow
At 11:15 PM -0700 7/18/05, Gary Clemans-Gibbon wrote: Thanks David, I just tried that line but it seems to be the same or if anything it seems even slower. I missed the start of this thread, but make sure that you do not have a duplex-mismatch with your ethernet cards. You want to be sure your server and your client(s) have the same duplex setting as whatever they are connected to (usually some switch). At one point we had excruciating performance problems with our samba server, and eventually this turned out to be a duplex mismatch between the server and the switch it was connected to. The switch was set to 100-Mb-full/duplex (which is to say, it was *not* doing autonegotiation), and the network interface on the server was set to autonegotiate. The server ended up with the wrong answer for duplex... -- Garance Alistair Drosehn= [EMAIL PROTECTED] Senior Systems Programmer or [EMAIL PROTECTED] Rensselaer Polytechnic Instituteor [EMAIL PROTECTED]
Re: Writes to samba server very, very slow
...on Tue, Jul 19, 2005 at 12:56:03AM -0700, Tim Hammerquist wrote: > I've seen this same phenomenon when copying to from my OSX Powerbook and > my fileserver (running both FreeBSD 5 and Gentoo Linux), with the OSX > acting as samba client. Just to drop this in again - at least here, writing from an OS X (10.2 or 10.3) system to a Samba 2 server is abysmally slow up to the point that OS X claims to have lost connection to the server, regardless of whether the Samba server runs on OpenBSD or Linux. Couldn't be bothered to find out the reason for that yet, though... Also didn't really try Samba 3 yet, as it didn't play nice with my OS/2 client last time... Alex.
Re: Writes to samba server very, very slow
On Tue, Jul 19, 2005 at 02:34:04PM +0200, Michael Hamerski wrote: > the FAQ which you refer to mentions 1M per 1G of storage, so that's not > really 1G of RAM for this system, is it? or is there a reason I'm missing? no...256M would in theory do it (assuming nothing bigger than around 200G in one partition, and fsck'd only one partition at a time), assuming your only goal was bringing the system back up. I'm assuming there is some USE for that huge disk system, that use will usually require some RAM. 200G is freaking huge. This machine probably shipped with an 8G disk and 128M of RAM, it got a major storage upgrade, and the rest was untouched. That's kinda like dropping a huge engine in your small car and assuming you have the ability to run with the pros. Not gonna happen. 1G might be way more than needed, but 256M is really still too tight for my comfort on a system of that size. > I am curious as I have a number of lower-end file-serving systems with > 200G - 500G and usually 256M RAM and have never been bit by a fsck > slowdown, even rebuilding raidframe parity is tolerable. Granted I have > partitions usually smaller than 50G and a system partition smaller than > 10G in most cases. Is this the difference? yes. The issue is, the larger the PARTITION, the more RAM it takes to fsck it. Supposedly, you can run multiple fsck's simultaniously. With multiple disks, there would be a definite advantage to doing so...with multiple big disks, big advantage...that moves the memory requirement up to near 512M. I don't believe RAIDframe adds a significant memory requirement based on disk size.
Re: Writes to samba server very, very slow
Marius Van Deventer - Umzimkulu wrote: Or you could disable apm0 and see if that helps. -Original Message- From: David Gwynne [mailto:[EMAIL PROTECTED] Sent: 19 July 2005 01:57 PM To: Gary Clemans-Gibbon Cc: misc@openbsd.org Subject: Re: Writes to samba server very, very slow From: "Gary Clemans-Gibbon" <[EMAIL PROTECTED]> Thanks for your reply Tim. If anything it makes me feel worse. I was hoping it was something easily fixed. I just tried transferring a 50 Mb file to the OBSD samba box from win using SCP. Again very slow writes but much faster reads. The 50 Mb file took about 7 mins to transfer to the OBSD box and about 30 seconds to read from the OBSD box. it looks like you're running a release 3.7 kernel. it is possible that your machine suffers from the halt-halt idle loop bug that has been fixed in stable. could you try upgrading your kernel to the stable branch and trying again? Hello, Ok, this is just a shot in the dark. If your slowness is across the network and protocol independant (ie: ftp, scp)... I used to see this alot when 100M NIC's first came out & the autodetection would not work properly. I think it was the cards would get confused with the full duplex/1/2 duplex setting, and would happen only when it was 2 100M devices... never when it was a 10<-->100 NIC This would affect Windows, Unix, etc... I used to just manually set the cards to the protocol I knew the switch was trying to do. This was YEARS ago, but who knows... but it was completely consistant that it was an order of magnitude slower only in one direction! Maybe this will help.. Cheers, Steve
Re: Writes to samba server very, very slow
* Michael Hamerski <[EMAIL PROTECTED]> [2005-07-19 14:57]: > I am curious as I have a number of lower-end file-serving systems with > 200G - 500G and usually 256M RAM and have never been bit by a fsck > slowdown, even rebuilding raidframe parity is tolerable. Granted I have > partitions usually smaller than 50G and a system partition smaller than > 10G in most cases. Is this the difference? yes. -- BS Web Services, http://www.bsws.de/ OpenBSD-based Webhosting, Mail Services, Managed Servers, ... Unix is very simple, but it takes a genius to understand the simplicity. (Dennis Ritchie)
Re: Writes to samba server very, very slow
Nick Holland wrote: Gary Clemans-Gibbon wrote: Hi All, I just built a OpenBSD 3.7 samba file server for my home lan. It's a P3 500, 128mb RAM, with a 2 gig IDE HDD for the OS and two x Maxtor 200 GB IDE drives for data. whoa. no where near enough RAM. Trip over the power cord, you will end up swapping during fsck. That will be most unpleasant. Assuming you have the swap available. (also assuming you have big partitions) 128M of RAM is large for a home firewall. 200G is a large amount of disk space, period. 2x200G is a huge amount of storage. Pair it up with some real RAM. In fact, if you are after balance, you will want about 1G of RAM, that won't be fun on that system, I suspect. Nick, the FAQ which you refer to mentions 1M per 1G of storage, so that's not really 1G of RAM for this system, is it? or is there a reason I'm missing? I am curious as I have a number of lower-end file-serving systems with 200G - 500G and usually 256M RAM and have never been bit by a fsck slowdown, even rebuilding raidframe parity is tolerable. Granted I have partitions usually smaller than 50G and a system partition smaller than 10G in most cases. Is this the difference? mike
Re: Writes to samba server very, very slow
Or you could disable apm0 and see if that helps. > -Original Message- > From: David Gwynne [mailto:[EMAIL PROTECTED] > Sent: 19 July 2005 01:57 PM > To: Gary Clemans-Gibbon > Cc: misc@openbsd.org > Subject: Re: Writes to samba server very, very slow > > > From: "Gary Clemans-Gibbon" <[EMAIL PROTECTED]> > > > Thanks for your reply Tim. If anything it makes me feel > worse. I was > > hoping it was something easily fixed. > > > > I just tried transferring a 50 Mb file to the OBSD samba > box from win > > using SCP. Again very slow writes but much faster reads. > The 50 Mb file > > took about 7 mins to transfer to the OBSD box and about 30 > seconds to read > > from the OBSD box. > > > > it looks like you're running a release 3.7 kernel. it is > possible that your > machine suffers from the halt-halt idle loop bug that has > been fixed in > stable. could you try upgrading your kernel to the stable > branch and trying > again?
Re: Writes to samba server very, very slow
From: "Gary Clemans-Gibbon" <[EMAIL PROTECTED]> Thanks for your reply Tim. If anything it makes me feel worse. I was hoping it was something easily fixed. I just tried transferring a 50 Mb file to the OBSD samba box from win using SCP. Again very slow writes but much faster reads. The 50 Mb file took about 7 mins to transfer to the OBSD box and about 30 seconds to read from the OBSD box. it looks like you're running a release 3.7 kernel. it is possible that your machine suffers from the halt-halt idle loop bug that has been fixed in stable. could you try upgrading your kernel to the stable branch and trying again?
Re: Writes to samba server very, very slow
Gary Clemans-Gibbon wrote: > Hi All, > > I just built a OpenBSD 3.7 samba file server for my home lan. It's a P3 > 500, 128mb RAM, with a 2 gig IDE HDD for the OS and two x Maxtor 200 GB > IDE drives for data. whoa. no where near enough RAM. Trip over the power cord, you will end up swapping during fsck. That will be most unpleasant. Assuming you have the swap available. (also assuming you have big partitions) 128M of RAM is large for a home firewall. 200G is a large amount of disk space, period. 2x200G is a huge amount of storage. Pair it up with some real RAM. In fact, if you are after balance, you will want about 1G of RAM, that won't be fun on that system, I suspect. http://www.openbsd.org/faq/faq14.html#LargeDrive I think you put a bullet in the project with that 2G "OS" drive, too. What were you planning on gaining by pairing up a 2G drive with a nice, brand-new 200G drive? (and it's not even a 2G, its a 1G...) > Everything is working fine except that when I copy files to the box from > a Windows XP box the transfers are very slow, like 9 minutes for a 48 Mb > file. Copying the same file back to the win box is quick - a couple of > seconds as you'd expect. Would you happen to be mirroring the drives with ccd(4)? That will hurt your write performance significantly. ... > > Here is dmesg > > OpenBSD 3.7 (GENERIC) #50: Sun Mar 20 00:01:57 MST 2005 > [EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC ...so not running RAIDframe... > wd0 at pciide0 channel 0 drive 0: > wd0: 16-sector PIO, LBA, 1032MB, 2114180 sectors > wd1 at pciide0 channel 0 drive 1: > wd1: 16-sector PIO, LBA48, 194481MB, 398297088 sectors > wd0(pciide0:0:0): using PIO mode 4, DMA mode 2 > wd1(pciide0:0:1): using PIO mode 4, Ultra-DMA mode 2 Mixing a 1G drive with a 200G drive on the same channel is really bad. I would have bet money that this config would not have worked for a few reasons: 1) small, old drive as master, big drive as slave 2) 200G drive on old IDE interface (pre-LBA48) 3) very very slow IDE drive with very fast IDE drive I'm amazed this thing boots. Anything beyond that is a gift. > wd2 at pciide0 channel 1 drive 0: > wd2: 16-sector PIO, LBA48, 194481MB, 398297088 sectors > atapiscsi0 at pciide0 channel 1 drive 1 > scsibus0 at atapiscsi0: 2 targets > cd0 at scsibus0 targ 0 lun 0: <, 52X24X52 CD-RW, 1.07> SCSI0 5/cdrom > removable > wd2(pciide0:1:0): using PIO mode 4, Ultra-DMA mode 2 > cd0(pciide0:1:1): using PIO mode 4, Ultra-DMA mode 2 and you sabbotage the secondary channel by mixing a HD with a CDROM. IF you really want that 1G drive, put it on the same channel with the CDROM, and put the two 200G drives on the same channel. Better design would be to put the two new, fast 200G drives on a new, fast PCI IDE card, and not use the on-board IDE interface. I'm amazed this works, I'm not surprised it is slow. Nick.
Re: Writes to samba server very, very slow
This *may* help. man mount softdep (FFS only.) Mount the file system using soft dependen- cies. Instead of metadata being written immediately, it is written in an ordered fashion to keep the on-disk state of the file system consistent. This results in significant speedups for file create/delete operations. This option will be ignored when using the -u flag and a file system is already mounted read/write. It requires option FFS_SOFTUPDATES to be enabled in the running ker- nel. There is a tradeoff between speed and safety. A rather large tradeoff I suspect ;) With any disk system, there is the question of what happens when the power fails. What is the speed when you copy the 48MB file locally? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Gary Clemans-Gibbon Sent: Tuesday, July 19, 2005 3:45 AM To: [EMAIL PROTECTED] Cc: misc@openbsd.org Subject: Re: Writes to samba server very, very slow Thanks for your reply Tim. If anything it makes me feel worse. I was hoping it was something easily fixed. I just tried transferring a 50 Mb file to the OBSD samba box from win using SCP. Again very slow writes but much faster reads. The 50 Mb file took about 7 mins to transfer to the OBSD box and about 30 seconds to read from the OBSD box. Perhaps this isn't a samba smb issue at all. My fstab... # cat /etc/fstab /dev/wd0a / ffs rw 1 1 /dev/wd1a /data1 ffs rw 1 2 /dev/wd2a /data2 ffs rw 1 2 same result with either data disk. I've been googling all evening and found many many forum posts with similar problems but no solutions. Some posts date back to 2002! If I have to go back to RH7.3 I'll be bummed. Especially as I spent ages setting up all my families accounts and softlinks for the data store. Waste of a day! Tim Hammerquist wrote: > Gary Clemans-Gibbon wrote: > >>David Gwynne wrote: >> >>>Gary Clemans-Gibbon wrote: >>> >>>>Everything is working fine except that when I copy files to the >>>>box from a Windows XP box the transfers are very slow, like >>>>9 minutes for a 48 Mb file. Copying the same file back to the win >>>>box is quick - a couple of seconds as you'd expect. >>> >>>I would suggest looking at the socket options parameter in >>>/etc/samba/ smb.conf. I have the following in my smb.conf and >>>transfer speeds seem to perform a lot better now: >>> >>>socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192 >> >>I just tried that line but it seems to be the same or if anything it >>seems even slower. > > > Gary, > > I've seen this same phenomenon when copying to from my OSX Powerbook and > my fileserver (running both FreeBSD 5 and Gentoo Linux), with the OSX > acting as samba client. > > The transfer speeds are not "slightly" slower, they are slower by orders > of magnitude, with normally 20sec transfers taking 10-20 minutes. > I watch the progress meter slowly incrementing at the rate of 32-64k/sec > over a 100bTX link. Does this sound like your issue? > > In my setup, I had limited success merely unmounting and remounting the > share; that worked maybe 50% of the time. Also, the rate seemed to be > normal more often if I had a simultaneous ssh connection between the > same two machines, even if the ssh connection were idle. I was not able > to find any consistently effective solution. > > After googling many times over several months, finding nothing more than > the same advice you got about TCP_NODELAY and the SO_*BUF settings > (which did not affect performance in my case either), I finally gave up, > switching to NFS and/or scp. > > For what it's worth, I haven't noticed this since I upgraded my > powerbook to OSX 10.4, so it might have something to do with the client > OS, network stack, or Samba version. > > I apologize for not having anything solid to recommend. But I wanted to > let you know that this *has* happened to others; you're not imagining > it. > > Tim Hammerquist > > > .
Re: Writes to samba server very, very slow
Gary Clemans-Gibbon wrote: > Thanks for your reply Tim. If anything it makes me feel worse. I was > hoping it was something easily fixed. > > I just tried transferring a 50 Mb file to the OBSD samba box from win > using SCP. Again very slow writes but much faster reads. The 50 Mb > file took about 7 mins to transfer to the OBSD box and about 30 > seconds to read from the OBSD box. > > Perhaps this isn't a samba smb issue at all. Indeed. Perhaps I misunderstood your issue. In my case, having a concurrent ssh connection between the same boxen seemed to increase the chance of having a good transfer speed via Samba. (This was in no way a reliable or even directly related effect.) However, scp transfer speeds, both ways, were always normal (read: good). Samba was the *only* protocol which suffered. With that in mind, I'm hoping your problem is indeed different from mine, and that you find a simple solution. If so, please post it to the list with "SOLVED" in the subject line. I'd be very interested in hearing it, and it would then be available in the archives. Cheers, Tim Hammerquist
Re: Writes to samba server very, very slow
Thanks for your reply Tim. If anything it makes me feel worse. I was hoping it was something easily fixed. I just tried transferring a 50 Mb file to the OBSD samba box from win using SCP. Again very slow writes but much faster reads. The 50 Mb file took about 7 mins to transfer to the OBSD box and about 30 seconds to read from the OBSD box. Perhaps this isn't a samba smb issue at all. My fstab... # cat /etc/fstab /dev/wd0a / ffs rw 1 1 /dev/wd1a /data1 ffs rw 1 2 /dev/wd2a /data2 ffs rw 1 2 same result with either data disk. I've been googling all evening and found many many forum posts with similar problems but no solutions. Some posts date back to 2002! If I have to go back to RH7.3 I'll be bummed. Especially as I spent ages setting up all my families accounts and softlinks for the data store. Waste of a day! Tim Hammerquist wrote: Gary Clemans-Gibbon wrote: David Gwynne wrote: Gary Clemans-Gibbon wrote: Everything is working fine except that when I copy files to the box from a Windows XP box the transfers are very slow, like 9 minutes for a 48 Mb file. Copying the same file back to the win box is quick - a couple of seconds as you'd expect. I would suggest looking at the socket options parameter in /etc/samba/ smb.conf. I have the following in my smb.conf and transfer speeds seem to perform a lot better now: socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192 I just tried that line but it seems to be the same or if anything it seems even slower. Gary, I've seen this same phenomenon when copying to from my OSX Powerbook and my fileserver (running both FreeBSD 5 and Gentoo Linux), with the OSX acting as samba client. The transfer speeds are not "slightly" slower, they are slower by orders of magnitude, with normally 20sec transfers taking 10-20 minutes. I watch the progress meter slowly incrementing at the rate of 32-64k/sec over a 100bTX link. Does this sound like your issue? In my setup, I had limited success merely unmounting and remounting the share; that worked maybe 50% of the time. Also, the rate seemed to be normal more often if I had a simultaneous ssh connection between the same two machines, even if the ssh connection were idle. I was not able to find any consistently effective solution. After googling many times over several months, finding nothing more than the same advice you got about TCP_NODELAY and the SO_*BUF settings (which did not affect performance in my case either), I finally gave up, switching to NFS and/or scp. For what it's worth, I haven't noticed this since I upgraded my powerbook to OSX 10.4, so it might have something to do with the client OS, network stack, or Samba version. I apologize for not having anything solid to recommend. But I wanted to let you know that this *has* happened to others; you're not imagining it. Tim Hammerquist .
Re: Writes to samba server very, very slow
Gary Clemans-Gibbon wrote: > David Gwynne wrote: > > Gary Clemans-Gibbon wrote: > > > Everything is working fine except that when I copy files to the > > > box from a Windows XP box the transfers are very slow, like > > > 9 minutes for a 48 Mb file. Copying the same file back to the win > > > box is quick - a couple of seconds as you'd expect. > > > > I would suggest looking at the socket options parameter in > > /etc/samba/ smb.conf. I have the following in my smb.conf and > > transfer speeds seem to perform a lot better now: > > > > socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192 > > I just tried that line but it seems to be the same or if anything it > seems even slower. Gary, I've seen this same phenomenon when copying to from my OSX Powerbook and my fileserver (running both FreeBSD 5 and Gentoo Linux), with the OSX acting as samba client. The transfer speeds are not "slightly" slower, they are slower by orders of magnitude, with normally 20sec transfers taking 10-20 minutes. I watch the progress meter slowly incrementing at the rate of 32-64k/sec over a 100bTX link. Does this sound like your issue? In my setup, I had limited success merely unmounting and remounting the share; that worked maybe 50% of the time. Also, the rate seemed to be normal more often if I had a simultaneous ssh connection between the same two machines, even if the ssh connection were idle. I was not able to find any consistently effective solution. After googling many times over several months, finding nothing more than the same advice you got about TCP_NODELAY and the SO_*BUF settings (which did not affect performance in my case either), I finally gave up, switching to NFS and/or scp. For what it's worth, I haven't noticed this since I upgraded my powerbook to OSX 10.4, so it might have something to do with the client OS, network stack, or Samba version. I apologize for not having anything solid to recommend. But I wanted to let you know that this *has* happened to others; you're not imagining it. Tim Hammerquist
Re: Writes to samba server very, very slow - more info
This is really puzzling me, please someone help me out. I've tried a couple of things. Firstly I swapped out the NIC for a different brand, no change. Writes god-awful slow, reads nice and zippy. I did some googling and tried some things. Here are the results of ifconfig -a netstat -in netstat -ss I don't know if the output from netstat -ss is normal or not there are duplicate packets, please let me know, $ ifconfig -a lo0: flags=8049 mtu 33224 inet 127.0.0.1 netmask 0xff00 inet6 ::1 prefixlen 128 inet6 fe80::1%lo0 prefixlen 64 scopeid 0x5 dc0: flags=8843 mtu 1500 address: 00:50:bf:9c:62:e4 media: Ethernet autoselect (100baseTX full-duplex) status: active inet 192.168.20.3 netmask 0xff00 broadcast 192.168.20.255 inet6 fe80::250:bfff:fe9c:62e4%dc0 prefixlen 64 scopeid 0x1 pflog0: flags=0<> mtu 33224 pfsync0: flags=0<> mtu 2020 enc0: flags=0<> mtu 1536 $ netstat -in NameMtu Network Address Ipkts IerrsOpkts Oerrs Colls lo0 332240 00 0 0 lo0 33224 127/8 127.0.0.10 00 0 0 lo0 33224 ::1/128 ::1 0 00 0 0 lo0 33224 fe80::%lo0/ fe80::1%lo0 0 00 0 0 dc0 150000:50:bf:9c:62:e424956 040617 0 0 dc0 1500 192.168.20/ 192.168.20.3 24956 040617 0 0 dc0 1500 fe80::%dc0/ fe80::250:bfff:fe24956 040617 0 0 pflog0* 332240 00 0 0 pfsync0 2020 0 00 0 0 enc0* 1536 0 00 0 0 $ netstat -ss ip: 24911 total packets received 16 bad header checksums 186 with data size < data length 24709 packets for this host 40602 packets sent from this host icmp: igmp: ipencap: tcp: 40579 packets sent 38744 data packets (50822440 bytes) 1 data packet (251 bytes) retransmitted 3 fast retransmitted packets 1354 ack-only packets (6063 delayed) 480 window update packets 24660 packets received 21738 acks (for 50819474 bytes) 58 duplicate acks 7062 packets (3128272 bytes) received in-sequence 52 completely duplicate packets (74096 bytes) 46 packets with some duplicate data (41032 bytes duplicated) 296 out-of-order packets (428428 bytes) 35 discarded for bad checksums 2 connection accepts 2 connections established (including accepts) 1 connection closed (including 0 drops) 12500 segments updated rtt (of 12502 attempts) 1 retransmit timeout 14841 correct ACK header predictions 2478 correct data packet header predictions 7 PCB cache misses cwr by fastrecovery: 3 cwr by timeout: 1 2 SYN cache entries added 2 completed udp: 58 datagrams received 1 broadcast/multicast datagram dropped due to no socket 57 delivered 30 datagrams output 1 missed PCB cache esp: ah: etherip: ipcomp: carp: pfsync: ip6: 7 packets sent from this host Mbuf statistics: icmp6: Output packet histogram: multicast listener report: 6 neighbor solicitation: 1 Histogram of error messages to be generated: pim6: rip6: thanks Gary Gary Clemans-Gibbon wrote: Thanks David, I just tried that line but it seems to be the same or if anything it seems even slower. Gary David Gwynne wrote: I would suggest looking at the socket options parameter in /etc/samba/ smb.conf. I have the following in my smb.conf and transfer speeds seem to perform a lot better now: socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192 On 19/07/2005, at 2:03 PM, Gary Clemans-Gibbon wrote: Hi All, I just built a OpenBSD 3.7 samba file server for my home lan. It's a P3 500, 128mb RAM, with a 2 gig IDE HDD for the OS and two x Maxtor 200 GB IDE drives for data. Everything is working fine except that when I copy files to the box from a Windows XP box the transfers are very slow, like 9 minutes for a 48 Mb file. Copying the same file back to the win box is quick - a couple of seconds as you'd expect. Please forgive me if I don't provide all the needed info here or if I didn't run any obvious checks. Please indicate what info is needed and how to get it and I'll repost it. This same hardware was previously running RH7.3 with samba and worked fine. I've tried a different ethernet cable and a different port on my switch too. many thanks in advance, Gary Here is /etc/samba/smb.conf global section.. [glo
Re: Writes to samba server very, very slow
Thanks David,
I just tried that line but it seems to be the same or if anything it
seems even slower.
Gary
David Gwynne wrote:
I would suggest looking at the socket options parameter in /etc/samba/
smb.conf. I have the following in my smb.conf and transfer speeds seem
to perform a lot better now:
socket options = TCP_NODELAY SO_RCVBUF=8192 SO_SNDBUF=8192
On 19/07/2005, at 2:03 PM, Gary Clemans-Gibbon wrote:
Hi All,
I just built a OpenBSD 3.7 samba file server for my home lan. It's a P3
500, 128mb RAM, with a 2 gig IDE HDD for the OS and two x Maxtor 200
GB IDE drives for data.
Everything is working fine except that when I copy files to the box from
a Windows XP box the transfers are very slow, like 9 minutes for a 48 Mb
file. Copying the same file back to the win box is quick - a couple
of seconds as you'd expect.
Please forgive me if I don't provide all the needed info here or if I
didn't run any obvious checks. Please indicate what info is needed
and how to get it and I'll repost it.
This same hardware was previously running RH7.3 with samba and worked
fine. I've tried a different ethernet cable and a different port on
my switch too.
many thanks in advance,
Gary
Here is /etc/samba/smb.conf global section..
[global]
workgroup = myworkgroup
server string = My Samba Server
hosts allow = 192.168.20. 127.0.0.1
log file = /var/log/smbd.%m
security = user
socket options = TCP_NODELAY IPTOS_LOWDELAY
read raw = yes
write raw = yes
Here is dmesg
OpenBSD 3.7 (GENERIC) #50: Sun Mar 20 00:01:57 MST 2005
[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel Pentium III ("GenuineIntel" 686-class, 512KB L2 cache)
501 MHz
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER
,MMX,FXSR,SSE
cpu0: disabling processor serial number
real mem = 133787648 (130652K)
avail mem = 115580928 (112872K)
using 1658 buffers containing 6791168 bytes (6632K) of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @
0xf0520
apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: APM engage (device 1): power management disabled (1)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev
0x00)
pcibios0: PCI bus #1 is the last bus
bios0: ROM list: 0xc/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER
,MMX,FXSR,SSE
cpu0: disabling processor serial number
real mem = 133787648 (130652K)
avail mem = 115580928 (112872K)
using 1658 buffers containing 6791168 bytes (6632K) of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @
0xf0520
apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: APM engage (device 1): power management disabled (1)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev
0x00)
pcibios0: PCI bus #1 is the last bus
bios0: ROM list: 0xc/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
pci1 at ppb0 bus 1
vga1 at pci1 dev 0 function 0 "Nvidia Riva TNT" rev 0x04
wsdisplay0 at vga1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
pcib0 at pci0 dev 4 function 0 "Intel 82371AB PIIX4 ISA" rev 0x02
pciide0 at pci0 dev 4 function 1 "Intel 82371AB IDE" rev 0x01: DMA,
channel 0 wired to compatibility, channel 1 wired to compatibility
wd0 at pciide0 channel 0 drive 0:
wd0: 16-sector PIO, LBA, 1032MB, 2114180 sectors
wd1 at pciide0 channel 0 drive 1:
wd1: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
wd0(pciide0:0:0): using PIO mode 4, DMA mode 2
wd1(pciide0:0:1): using PIO mode 4, Ultra-DMA mode 2
wd2 at pciide0 channel 1 drive 0:
wd2: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
atapiscsi0 at pciide0 channel 1 drive 1
scsibus0 at atapiscsi0: 2 targets
cd0 at scsibus0 targ 0 lun 0: <, 52X24X52 CD-RW, 1.07> SCSI0 5/ cdrom
removable
wd2(pciide0:1:0): using PIO mode 4, Ultra-DMA mode 2
cd0(pciide0:1:1): using PIO mode 4, Ultra-DMA mode 2
uhci0 at pci0 dev 4 function 2 "Intel 82371AB USB" rev 0x01: irq 9
usb0 at uhci0: USB revision 1.0
uhub0 at usb0
uhub0: Intel UHCI root hub, class 9/0, rev 1.00/1.00, addr 1
uh
Writes to samba server very, very slow
Hi All,
I just built a OpenBSD 3.7 samba file server for my home lan. It's a P3
500, 128mb RAM, with a 2 gig IDE HDD for the OS and two x Maxtor 200 GB
IDE drives for data.
Everything is working fine except that when I copy files to the box from
a Windows XP box the transfers are very slow, like 9 minutes for a 48 Mb
file. Copying the same file back to the win box is quick - a couple of
seconds as you'd expect.
Please forgive me if I don't provide all the needed info here or if I
didn't run any obvious checks. Please indicate what info is needed and
how to get it and I'll repost it.
This same hardware was previously running RH7.3 with samba and worked
fine. I've tried a different ethernet cable and a different port on my
switch too.
many thanks in advance,
Gary
Here is /etc/samba/smb.conf global section..
[global]
workgroup = myworkgroup
server string = My Samba Server
hosts allow = 192.168.20. 127.0.0.1
log file = /var/log/smbd.%m
security = user
socket options = TCP_NODELAY IPTOS_LOWDELAY
read raw = yes
write raw = yes
Here is dmesg
OpenBSD 3.7 (GENERIC) #50: Sun Mar 20 00:01:57 MST 2005
[EMAIL PROTECTED]:/usr/src/sys/arch/i386/compile/GENERIC
cpu0: Intel Pentium III ("GenuineIntel" 686-class, 512KB L2 cache) 501 MHz
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER,MMX,FXSR,SSE
cpu0: disabling processor serial number
real mem = 133787648 (130652K)
avail mem = 115580928 (112872K)
using 1658 buffers containing 6791168 bytes (6632K) of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @ 0xf0520
apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: APM engage (device 1): power management disabled (1)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev 0x00)
pcibios0: PCI bus #1 is the last bus
bios0: ROM list: 0xc/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
cpu0:
FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,SER,MMX,FXSR,SSE
cpu0: disabling processor serial number
real mem = 133787648 (130652K)
avail mem = 115580928 (112872K)
using 1658 buffers containing 6791168 bytes (6632K) of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(06) BIOS, date 03/03/00, BIOS32 rev. 0 @ 0xf0520
apm0 at bios0: Power Management spec V1.2 (BIOS mgmt disabled)
apm0: APM power management enable: unrecognized device ID (9)
apm0: APM engage (device 1): power management disabled (1)
apm0: AC on, battery charge unknown
pcibios0 at bios0: rev 2.1 @ 0xf/0xd92
pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xf0d10/128 (6 entries)
pcibios0: PCI Interrupt Router at 000:04:0 ("Intel 82371FB ISA" rev 0x00)
pcibios0: PCI bus #1 is the last bus
bios0: ROM list: 0xc/0x8000
cpu0 at mainbus0
pci0 at mainbus0 bus 0: configuration mode 1 (no bios)
pchb0 at pci0 dev 0 function 0 "Intel 82443BX AGP" rev 0x03
ppb0 at pci0 dev 1 function 0 "Intel 82443BX AGP" rev 0x03
pci1 at ppb0 bus 1
vga1 at pci1 dev 0 function 0 "Nvidia Riva TNT" rev 0x04
wsdisplay0 at vga1: console (80x25, vt100 emulation)
wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
pcib0 at pci0 dev 4 function 0 "Intel 82371AB PIIX4 ISA" rev 0x02
pciide0 at pci0 dev 4 function 1 "Intel 82371AB IDE" rev 0x01: DMA,
channel 0 wired to compatibility, channel 1 wired to compatibility
wd0 at pciide0 channel 0 drive 0:
wd0: 16-sector PIO, LBA, 1032MB, 2114180 sectors
wd1 at pciide0 channel 0 drive 1:
wd1: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
wd0(pciide0:0:0): using PIO mode 4, DMA mode 2
wd1(pciide0:0:1): using PIO mode 4, Ultra-DMA mode 2
wd2 at pciide0 channel 1 drive 0:
wd2: 16-sector PIO, LBA48, 194481MB, 398297088 sectors
atapiscsi0 at pciide0 channel 1 drive 1
scsibus0 at atapiscsi0: 2 targets
cd0 at scsibus0 targ 0 lun 0: <, 52X24X52 CD-RW, 1.07> SCSI0 5/cdrom
removable
wd2(pciide0:1:0): using PIO mode 4, Ultra-DMA mode 2
cd0(pciide0:1:1): using PIO mode 4, Ultra-DMA mode 2
uhci0 at pci0 dev 4 function 2 "Intel 82371AB USB" rev 0x01: irq 9
usb0 at uhci0: USB revision 1.0
uhub0 at usb0
uhub0: Intel UHCI root hub, class 9/0, rev 1.00/1.00, addr 1
uhub0: 2 ports with 2 removable, self powered
"Intel 82371AB Power Mgmt" rev 0x02 at pci0 dev 4 function 3 not configured
dc0 at pci0 dev 9 function 0 "ADMtek AN983" rev 0x11: irq 9, address
00:50:bf:fe:83:bc
ukphy0 at dc0 phy 1: Generic IEEE 802.3u media interface
ukphy0: OUI 0x000749, model 0x0001, rev. 1
isa0 at pcib0
isadma0 at isa0
pckbc0 at isa0 port 0x60/5
pckbd0 at pckbc0 (kbd slot)
pckbc0: using irq 1 for kbd slot
wskbd0 at pckbd0 (mux 1 igno
