Date: Sun, 18 Dec 2005 02:32:52 -0800 From: "Ted Mittelstaedt" <tedm@toybox.placo.com> To: "Sasa Stupar" <sasa@stupar.homelinux.net>, <freebsd-questions@freebsd.org> Subject: RE: Polling For 100 mbps Connections? (Was Re: Freebsd Theme Song) Message-ID: <LOBBIFDAGNMAMLGJJCKNMEBDFDAA.tedm@toybox.placo.com> In-Reply-To: <6F4719B836FB9479D8754136@[192.168.10.249]>
next in thread | previous in thread | raw e-mail | index | archive | help
>-----Original Message----- >From: owner-freebsd-questions@freebsd.org >[mailto:owner-freebsd-questions@freebsd.org]On Behalf Of Sasa Stupar >Sent: Sunday, December 18, 2005 2:21 AM >To: Ted Mittelstaedt; freebsd-questions@freebsd.org >Subject: RE: Polling For 100 mbps Connections? (Was Re: Freebsd >Theme Song) > > > > >--On 18. december 2005 1:33 -0800 Ted Mittelstaedt ><tedm@toybox.placo.com> >wrote: > >> >> >>> -----Original Message----- >>> From: Sasa Stupar [mailto:sasa@stupar.homelinux.net] >>> Sent: Friday, December 16, 2005 5:25 AM >>> To: Ted Mittelstaedt; danial_thom@yahoo.com; Drew Tomlinson >>> Cc: freebsd-questions@freebsd.org >>> Subject: RE: Polling For 100 mbps Connections? (Was Re: Freebsd >>> Theme Song) >>> >>> >>> >>> >>> --On 16. december 2005 3:36 -0800 Ted Mittelstaedt >>> <tedm@toybox.placo.com> >>> wrote: >>> >>>> >>>> >>>>> -----Original Message----- >>>>> From: Sasa Stupar [mailto:sasa@stupar.homelinux.net] >>>>> Sent: Thursday, December 15, 2005 12:34 AM >>>>> To: Ted Mittelstaedt; danial_thom@yahoo.com; Drew Tomlinson >>>>> Cc: freebsd-questions@freebsd.org >>>>> Subject: RE: Polling For 100 mbps Connections? (Was Re: Freebsd >>>>> Theme Song) >>>>> >>>>>> >>>>>> Ted >>>>> >>>>> Hmmm, here is test with iperf what I have done with and >>> without polling: >>>>> ************** >>>>> ------------------------------------------------------------ >>>>> Client connecting to 192.168.1.200, TCP port 5001 >>>>> TCP window size: 8.00 KByte (default) >>>>> ------------------------------------------------------------ >>>>> [1816] local 192.168.10.249 port 1088 connected with >>>>> 192.168.1.200 port 5001 >>>>> [ ID] Interval Transfer Bandwidth >>>>> [1816] 0.0-10.0 sec 108 MBytes 90.1 Mbits/sec >>>>> >>>>> This is when I use Device polling option on m0n0. >>>>> >>>>> If I disable this option then my transfer is worse: >>>>> ------------------------------------------------------------ >>>>> Client connecting to 192.168.1.200, TCP port 5001 >>>>> TCP window size: 8.00 KByte (default) >>>>> ------------------------------------------------------------ >>>>> [1816] local 192.168.10.249 port 1086 connected with >>>>> 192.168.1.200 port 5001 >>>>> [ ID] Interval Transfer Bandwidth >>>>> [1816] 0.0-10.0 sec 69.7 MBytes 58.4 Mbits/sec >>>>> *************** >>>>> >>>>> BTW: my router is m0n0wall (FBSD 4.11). >>>>> >>>> >>>> what are the cpu speeds and operating systems of all devices >>>> in the packet path, what is the make and model of switchs in >>>> use, provide dmesg output of the bsd box, a network diagram >>>> of the setup, etc. etc. etc. >>>> >>>> The above test results are not replicatable and thus, worthless. >>>> Useful test results would allow a reader to build an exact >>>> duplicate of your setup, config it identically, and get identical >>>> results. >>>> >>>> Ted >>>> >>> >>> OK. The server (192.168.1.200) is FreeBSD 5.4 with Duron 900 >and 3C905C >> >> The 3com 3c905 is not a very good card under FreeBSD the driver was >> written >> without support from 3com and is shakey on a lot of hardware. I would >> say >> there's a big question that your server is actually saturating the >> ethernet. >> Probably that is why your only getting 90Mbt. >> >>> NIC; router is m0n0wall (FreeBSD 4.11) with three Intel >>> Pro/100S Nics and >>> Celeron 433; The user computer (192.168.10.249) is Celeron 2400 >>> with winxp >>> and integrated NIC Realtek 8139 series. Switch is CNET CNSH-1600. >>> >> >> Once again, the winxp+realtek 8139 is not a particularly >steller combo, >> I would question that this system could saturate the ethernet, either. >> >>> Diagram: <http://me.homelinux.net/network.pdf>; >>> >>> dmesg from the router: >>> ---------------- >>> $ dmesg >>> Copyright (c) 1992-2005 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 4.11-RELEASE-p11 #0: Wed Sep 7 13:49:09 CEST 2005 >>> root@fb411.neon1.net:/usr/src/sys/compile/M0N0WALL_GENERIC >>> Timecounter "i8254" frequency 1193182 Hz >>> CPU: Pentium II/Pentium II Xeon/Celeron (434.32-MHz 686-class CPU) >>> Origin = "GenuineIntel" Id = 0x665 Stepping = 5 >>> >>> Features=0x183f9ff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,P >>> GE,MCA,CMOV,PAT,PSE36,MMX,FXSR> >>> real memory = 201326592 (196608K bytes) >>> avail memory = 179142656 (174944K bytes) >>> Preloaded elf kernel "kernel" at 0xc1006000. >>> Preloaded mfs_root "/mfsroot" at 0xc100609c. >>> Pentium Pro MTRR support enabled >>> md0: Preloaded image </mfsroot> 11534336 bytes at 0xc0504d9c >>> md1: Malloc disk >>> Using $PIR table, 8 entries at 0xc00fdef0 >>> npx0: <math processor> on motherboard >>> npx0: INT 16 interface >>> pcib0: <Intel 82443BX (440 BX) host to PCI bridge> on motherboard >>> pci0: <PCI bus> on pcib0 >>> pcib1: <Intel 82443BX (440 BX) PCI-PCI (AGP) bridge> at device >>> 1.0 on pci0 >>> pci1: <PCI bus> on pcib1 >>> 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 >>> uhci0: <Intel 82371AB/EB (PIIX4) USB controller> port >>> 0xd000-0xd01f 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 >>> chip1: <Intel 82371AB Power management controller> port >>> 0x5000-0x500f at >>> device 7.3 on pci0 >>> pci0: <unknown card> (vendor=0x1274, dev=0x1371) at 8.0 irq 11 >>> fxp0: <Intel 82550 Pro/100 Ethernet> port 0xd800-0xd83f mem >>> 0xd0400000-0xd041ffff,0xd0460000-0xd0460fff irq 10 at device >>> 15.0 on pci0 >>> fxp0: Ethernet address 00:02:b3:62:f6:06 >>> inphy0: <i82555 10/100 media interface> on miibus0 >>> inphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto >>> fxp1: <Intel 82550 Pro/100 Ethernet> port 0xdc00-0xdc3f mem >>> 0xd0420000-0xd043ffff,0xd0462000-0xd0462fff irq 12 at device >>> 16.0 on pci0 >>> fxp1: Ethernet address 00:02:b3:9c:2a:16 >>> inphy1: <i82555 10/100 media interface> on miibus1 >>> inphy1: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto >>> fxp2: <Intel 82550 Pro/100 Ethernet> port 0xe000-0xe03f mem >>> 0xd0440000-0xd045ffff,0xd0461000-0xd0461fff irq 7 at device >19.0 on pci0 >>> fxp2: Ethernet address 00:02:b3:8c:e4:f6 >>> inphy2: <i82555 10/100 media interface> on miibus2 >>> inphy2: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto >>> pmtimer0 on isa0 >>> 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 >>> sio0 at port 0x3f8-0x3ff irq 4 flags 0x10 on isa0 >>> sio0: type 16550A, console >>> sio1: configured irq 3 not in bitmap of probed irqs 0 >>> BRIDGE 020214 loaded >>> IPsec: Initialized Security Association Processing. >>> IP Filter: v3.4.35 initialized. Default = block all, >Logging = enabled >>> ad0: 3098MB <WDC AC33200L> [6296/16/63] at ata0-master PIO4 >>> acd0: CDROM <LITE-ON CD-ROM LTN-527T> at ata1-master PIO4 >>> Mounting root from ufs:/dev/md0c >>> fxp1: Microcode loaded, int_delay: 1000 usec bundle_max: 6 >>> fxp0: Microcode loaded, int_delay: 1000 usec bundle_max: 6 >>> fxp2: Microcode loaded, int_delay: 1000 usec bundle_max: 6 >>> ata0: resetting devices .. done >>> ------------- >>> >>> If you need more just ask for it. You don't need to be angry. Peace. >>> >> >> OK, next question: >> >> ftp transfer like this uses large packets, rerun the test with ping -f >> with different ping packet sizes, post the results. >> >> Remember, routers have to deal with many sized packets. >> >> Ted >> > >Interesting. I have tested like you've said and I could ping >with packet >size 1450 bytes. Everything bigger is telling that "packet must be >fragmented but DF is set up". This is of course pinging from winxp to >server. That is normal since under winxp ping sets the DF bit I believe. The larger packets are not what matters, the smaller packets are more interesting. I find it hard to believe your getting the same throughput with flood pinging with 56 byte packets. Ted
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?LOBBIFDAGNMAMLGJJCKNMEBDFDAA.tedm>