NASLite Network Attached Storage

Hi everybody,

Before posting, I have tested and played around with different solutions but I can't resolve the problem any further. It really seems a NL related problem. The main problem is that the copy speeds change without me doing anything...

I have 3 XP PRO SP2 computers that will use NL2-USB server. I have built most of the computers for speed and they are recent computers. All have 1MB RAM, dual channel on the NL server and one of the XPs. 3 SATA2 drives on the NL server.

As you'll see below on the SysLog, the NeatGear GA311 is the only one to use IRQ 7, in the BIOS everything is shut down except USB (set to 1.1 as one member said (on the forum) that with the same MoBo (ASUS M2N-MX) he solved the freeze at boot with this option), the on board graphic chip and ACPI 3.0.

Computers 1 to 3 are XPs, server is NL. I have a modem-router (100MHz)-DHCP "IP server" connected to ADSL, LAN printer and a LAN web cam and to a D-Link Gigabit Switch DGS 1005D. All 4 computers connected to the GB switch with high quality Cat5 cable. All the connexions are on "auto negotiate" and connect systematically at 1 GB speed. IP numbers of the computers (except NL server) are not fix. I also have a laptop connected to the Modem Router by 54MHz WiFi but this computer is not used for the LAN but mainly for internet access and mails.

NL is connected with a drive letter (N:) from the XP Pro machines. The connexion is to \\Serveur\Disk-0 (or disk 2). I see no difference if the connexion is done to \\Serveur\Shares\Disk-0. Sometimes there is a speed difference if I copy to the "LAN favorites" (Favoris réseau in French) compared to using the "connected" drive letter, copy to the drive letter being a bit faster.

The read speeds from the NL server is very good and around 26 to 38 000 kB/s on all computers and always.

FTP reads AND writes are always very fast too (using CuteFTP Pro 7).

Read AND write speeds between XP Pro machines are ALWAYS good (15 to 28000 KB/s)

The write to NL server is the problematic issue. This morning, clean boot for everyone and I start testing with a file of >100 MB. From computer 3 to NL: 25000kB/S copy speed (I think "Great, the proble is solved" but...) About the same speed from computer 1 to NL. But only 400 kB/s from the computer 2 to NL. I test again with computer 3 and 1. The speed is not so slow but half the speed of the first copy. Then when I test some time later, it goes up to 36000. Then goes down to 20000 or 9000kB/s. Here is when I start loosing the logic of it . Computer 2 is often around the 400 kb/s but sometimes goes up to 10 000 kB/s. Yesterday, computer 1 was around the 400 - 500 kB/s speed all the time.

There are no collisions or other LAN problems noted.

As with all intermittent problems, this one is hard to solve and I don't know where to look for. I didn't have time to install FreeNAS or Ubuntu Server on the actual NL server to see how they compared to NL but when there was XP Pro installed on it, I didn't have any write speed problems with the computer.

My deduction for the moment is that there is a LAN problem related to NL in SMB/CIFS transfers. I'm new to Linux so I can't guess where.

Any help or ideas are welcomed

TIA for your time and help.

Robby

LAN device details

* eth0: Identified chip type is 'RTL8169s/8110s'.
* eth0: RTL8169 at 0xf8812c00, 00:18:4d:e9:cf:29, IRQ 7
* eth0: Auto-negotiation Enabled.
* eth0: 1000Mbps Full-duplex operation.
* eth0 Link encap:Ethernet HWaddr 00:18:4D:E9:CF:29
* inet addr:192.168.1.40 Bcast:192.168.1.255 Mask:255.255.255.0
* UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
* RX packets:1596014 errors:0 dropped:0 overruns:0 frame:0
* TX packets:1146584 errors:0 dropped:0 overruns:0 carrier:0
* collisions:0 txqueuelen:1000
* RX bytes:2068970462 (1.9 GiB) TX bytes:765009199 (729.5 MiB)
* Interrupt:7 Base address:0x2c00


SysLog

* Aug 1 05:57:30 syslog.info syslogd started: BusyBox v1.01 (2007.03.15-22:03+0000)
* Aug 1 05:57:30 daemon.info init: ^MStarting pid 2868, console /dev/null: '/sbin/klogd'
* Aug 1 05:57:30 user.notice kernel: klogd started: BusyBox v1.01 (2007.03.15-22:03+0000)
* Aug 1 05:57:30 user.warn kernel: Linux version 2.4.34.NASLite (root@tzt) (gcc version 3.4.6) #1 Wed Apr 11 22:39:05 EDT 2007
* Aug 1 05:57:30 user.info kernel: BIOS-provided physical RAM map:
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 0000000000000000 - 000000000009fc00 (usable)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 000000000009fc00 - 00000000000a0000 (reserved)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 00000000000e0000 - 0000000000100000 (reserved)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 0000000000100000 - 000000003efc0000 (usable)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 000000003efc0000 - 000000003efce000 (ACPI data)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 000000003efce000 - 000000003eff0000 (ACPI NVS)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 000000003eff0000 - 000000003f000000 (reserved)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 00000000fec00000 - 00000000fec01000 (reserved)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 00000000fee00000 - 00000000fef00000 (reserved)
* Aug 1 05:57:30 user.warn kernel: BIOS-e820: 00000000ff780000 - 0000000100000000 (reserved)
* Aug 1 05:57:30 user.warn kernel: Warning only 896MB will be used.
* Aug 1 05:57:30 user.warn kernel: Use a HIGHMEM enabled kernel.
* Aug 1 05:57:30 user.notice kernel: 896MB LOWMEM available.
* Aug 1 05:57:30 user.warn kernel: On node 0 totalpages: 229376
* Aug 1 05:57:30 user.warn kernel: zone(0): 4096 pages.
* Aug 1 05:57:30 user.warn kernel: zone(1): 225280 pages.
* Aug 1 05:57:30 user.warn kernel: zone(2): 0 pages.
* Aug 1 05:57:30 user.info kernel: ACPI: RSDP (v002 ACPIAM ) @ 0x000fb840
* Aug 1 05:57:30 user.info kernel: ACPI: XSDT (v001 _ASUS_ Notebook 0x04000730 MSFT 0x00000097) @ 0x3efc0100
* Aug 1 05:57:30 user.info kernel: ACPI: FADT (v003 A_M_I_ OEMFACP 0x04000730 MSFT 0x00000097) @ 0x3efc0290
* Aug 1 05:57:30 user.info kernel: ACPI: MADT (v001 A_M_I_ OEMAPIC 0x04000730 MSFT 0x00000097) @ 0x3efc0390
* Aug 1 05:57:30 user.info kernel: ACPI: MCFG (v001 A_M_I_ OEMMCFG 0x04000730 MSFT 0x00000097) @ 0x3efc0400
* Aug 1 05:57:30 user.info kernel: ACPI: SLIC (v001 _ASUS_ Notebook 0x04000730 MSFT 0x00000097) @ 0x3efc0440
* Aug 1 05:57:30 user.info kernel: ACPI: OEMB (v001 A_M_I_ AMI_OEM 0x04000730 MSFT 0x00000097) @ 0x3efce040
* Aug 1 05:57:30 user.info kernel: ACPI: SSDT (v001 A_M_I_ POWERNOW 0x00000001 AMD 0x00000001) @ 0x3efc5310
* Aug 1 05:57:30 user.info kernel: ACPI: DSDT (v001 A0557 A0557000 0x00000000 INTL 0x20051117) @ 0x00000000
* Aug 1 05:57:30 user.warn kernel: Kernel command line: rw root=/dev/ram0 initrd=NASLite.02 quiet BOOT_IMAGE=naslite.01
* Aug 1 05:57:30 user.info kernel: Initializing CPU#0
* Aug 1 05:57:30 user.warn kernel: Detected 2009.160 MHz processor.
* Aug 1 05:57:30 user.warn kernel: Console: colour VGA+ 80x25
* Aug 1 05:57:30 user.warn kernel: Calibrating delay loop. 4010.80 BogoMIPS
* Aug 1 05:57:30 user.info kernel: Memory: 901996k/917504k available (2182k kernel code, 15120k reserved, 629k data, 536k init, 0k highmem)
* Aug 1 05:57:30 user.info kernel: Dentry cache hash table entries: 131072 (order: 8, 1048576 bytes)
* Aug 1 05:57:30 user.info kernel: Inode cache hash table entries: 65536 (order: 7, 524288 bytes)
* Aug 1 05:57:30 user.info kernel: Mount cache hash table entries: 512 (order: 0, 4096 bytes)
* Aug 1 05:57:30 user.info kernel: Buffer cache hash table entries: 65536 (order: 6, 262144 bytes)
* Aug 1 05:57:30 user.warn kernel: Page-cache hash table entries: 262144 (order: 8, 1048576 bytes)
* Aug 1 05:57:30 user.info kernel: CPU: L1 I Cache: 64K (64 bytes/line), D cache 64K (64 bytes/line)
* Aug 1 05:57:30 user.info kernel: CPU: L2 Cache: 512K (64 bytes/line)
* Aug 1 05:57:30 user.debug kernel: CPU: After generic, caps: 078bfbff ebd3fbff 00000000 00000000
* Aug 1 05:57:30 user.debug kernel: CPU: Common caps: 078bfbff ebd3fbff 00000000 00000000
* Aug 1 05:57:30 user.warn kernel: CPU: AMD Athlon(tm) 64 Processor 3200+ stepping 02
* Aug 1 05:57:30 user.info kernel: Enabling fast FPU save and restore. done.
* Aug 1 05:57:30 user.info kernel: Enabling unmasked SIMD FPU exception support. done.
* Aug 1 05:57:30 user.info kernel: Checking 'hlt' instruction. OK.
* Aug 1 05:57:30 user.warn kernel: ACPI: IRQ9 SCI: Edge set to Level Trigger.
* Aug 1 05:57:30 user.warn kernel: POSIX conformance testing by UNIFIX
* Aug 1 05:57:30 user.info kernel: ACPI: Subsystem revision 20040326
* Aug 1 05:57:30 user.info kernel: PCI: PCI BIOS revision 3.00 entry at 0xf0031, last bus=4
* Aug 1 05:57:30 user.info kernel: PCI: Using configuration type 1
* Aug 1 05:57:30 user.info kernel: ACPI: Interpreter enabled
* Aug 1 05:57:30 user.info kernel: ACPI: Using PIC for interrupt routing
* Aug 1 05:57:30 user.info kernel: ACPI: System [ACPI] (supports S0 S1 S3 S4 S5)
* Aug 1 05:57:30 user.info kernel: ACPI: PCI Root Bridge [PCI0] (00:00)
* Aug 1 05:57:30 user.warn kernel: PCI: Probing PCI hardware (bus 00)
* Aug 1 05:57:30 user.warn kernel: Transparent bridge - PCI device 10de:03f3
* Aug 1 05:57:30 user.debug kernel: ACPI: PCI Interrupt Routing Table [\_SB_.PCI0._PRT]
* Aug 1 05:57:30 user.debug kernel: ACPI: PCI Interrupt Routing Table [\_SB_.PCI0.P0P1._PRT]
* Aug 1 05:57:30 user.debug kernel: ACPI: PCI Interrupt Routing Table [\_SB_.PCI0.P0P2._PRT]
* Aug 1 05:57:30 user.debug kernel: ACPI: PCI Interrupt Routing Table [\_SB_.PCI0.BR11._PRT]
* Aug 1 05:57:30 user.debug kernel: ACPI: PCI Interrupt Routing Table [\_SB_.PCI0.BR12._PRT]
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNKA] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNKB] (IRQs *7 10 11 14)
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNKC] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNKD] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNEA] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNEB] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNEC] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNED] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LUB0] (IRQs 7 *10 11 14)
* Aug 1 05:57:30 user.warn kernel: Link [LUB2] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LMAC] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LAZA] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LACI] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LMC9] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LSMB] (IRQs 7 10 *11 14)
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LPMU] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LSA0] (IRQs *15)
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LSA1] (IRQs *5)
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LATA] (IRQs 7 10 11 14) *0, disabled.
* Aug 1 05:57:30 user.info kernel: PCI: Probing PCI hardware
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LSMB] enabled at IRQ 11
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LUB0] enabled at IRQ 10
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LSA0] enabled at IRQ 15
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LSA1] enabled at IRQ 5
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LMC9] enabled at IRQ 11
* Aug 1 05:57:30 user.warn kernel: ACPI: PCI Interrupt Link [LNKB] enabled at IRQ 7
* Aug 1 05:57:30 user.info kernel: PCI: Using ACPI for IRQ routing
* Aug 1 05:57:30 user.info kernel: Linux NET4.0 for Linux 2.4
* Aug 1 05:57:30 user.info kernel: Based upon Swansea University Computer Society NET3.039
* Aug 1 05:57:30 user.warn kernel: Initializing RT netlink socket
* Aug 1 05:57:30 user.warn kernel: Starting kswapd
* Aug 1 05:57:30 user.info kernel: Journalled Block Device driver loaded
* Aug 1 05:57:30 user.info kernel: devfs: v1.12c (20020818) Richard Gooch (rgooch@atnf.csiro.au)
* Aug 1 05:57:30 user.info kernel: devfs: boot_options: 0x1
* Aug 1 05:57:30 user.info kernel: Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
* Aug 1 05:57:30 user.info kernel: ACPI: Power Button (FF) [PWRF]
* Aug 1 05:57:30 user.info kernel: ACPI: Processor [CPU1] (supports C1)
* Aug 1 05:57:30 user.info kernel: ACPI: Processor [CPU2] (supports C1)
* Aug 1 05:57:30 user.warn kernel: pty: 256 Unix98 ptys configured
* Aug 1 05:57:30 user.info kernel: Real Time Clock Driver v1.10f
* Aug 1 05:57:30 user.warn kernel: keyboard: Timeout - AT keyboard not present?(ed)
* Aug 1 05:57:30 user.warn kernel: keyboard: Timeout - AT keyboard not present?(f4)
* Aug 1 05:57:30 user.info kernel: FDC 0 is a post-1991 82077
* Aug 1 05:57:30 user.warn kernel: RAMDISK driver initialized: 16 RAM disks of 8192K size 1024 blocksize
* Aug 1 05:57:30 user.info kernel: loop: loaded (max 8 devices)
* Aug 1 05:57:30 user.info kernel: Intel(R) PRO/1000 Network Driver - version 5.7.6-k1
* Aug 1 05:57:30 user.info kernel: Copyright (c) 1999-2004 Intel Corporation.
* Aug 1 05:57:30 user.info kernel: dgrs: SW=$Id: dgrs.c,v 1.13 2000/06/06 04:07:00 rick Exp $ FW=Build 550 11/16/96 03:45:15
* Aug 1 05:57:30 user.warn kernel: FW Version=$Version$
* Aug 1 05:57:30 user.info kernel: pcnet32.c:v1.30h 06.24.2004 tsbogend@alpha.franken.de
* Aug 1 05:57:30 user.info kernel: ThunderLAN driver v1.15
* Aug 1 05:57:30 user.info kernel: TLAN: 0 devices installed, PCI: 0 EISA: 0
* Aug 1 05:57:30 user.info kernel: dmfe: Davicom DM9xxx net driver, version 1.36.4 (2002-01-17)
* Aug 1 05:57:30 user.info kernel: ns83820.c: National Semiconductor DP83820 10/100/1000 driver.
* Aug 1 05:57:30 user.warn kernel: sk98lin: No adapter found.
* Aug 1 05:57:30 user.info kernel: forcedeth.c: Reverse Engineered nForce ethernet driver. Version 0.50.
* Aug 1 05:57:30 user.info kernel: r8169 Gigabit Ethernet driver 1.2 loaded
* Aug 1 05:57:30 user.debug kernel: eth0: Identified chip type is 'RTL8169s/8110s'.
* Aug 1 05:57:30 user.info kernel: eth0: RTL8169 at 0xf8812c00, 00:18:4d:e9:cf:29, IRQ 7
* Aug 1 05:57:30 user.info kernel: eth0: Auto-negotiation Enabled.
* Aug 1 05:57:30 user.info kernel: eth0: 1000Mbps Full-duplex operation.
* Aug 1 05:57:30 user.info kernel: Uniform Multi-Platform E-IDE driver Revision: 7.00beta4-2.4
* Aug 1 05:57:30 user.info kernel: ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
* Aug 1 05:57:30 user.warn kernel: hda: LITE-ON DVDRW LH-20A1P, ATAPI CD/DVD-ROM drive
* Aug 1 05:57:30 user.warn kernel: ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
* Aug 1 05:57:30 user.info kernel: SCSI subsystem driver Revision: 1.00
* Aug 1 05:57:30 user.info kernel: Red Hat/Adaptec aacraid driver (1.1-3 Apr 11 2007 22:41:48)
* Aug 1 05:57:30 user.warn kernel: scsi: Detection failed (no card)
* Aug 1 05:57:30 user.notice kernel: megaraid: v2.10.10.1 (Release Date: Thu Jan 27 16:19:44 EDT 2005)
* Aug 1 05:57:30 user.warn kernel: GDT-HA: Storage RAID Controller Driver. Version: 3.04
* Aug 1 05:57:30 user.warn kernel: GDT-HA: Found 0 PCI Storage RAID Controllers
* Aug 1 05:57:30 user.warn kernel: 3ware Storage Controller device driver for Linux v1.02.00.037.
* Aug 1 05:57:30 user.warn kernel: 3w-xxxx: No cards found.
* Aug 1 05:57:30 user.debug kernel: libata version 1.20 loaded.
* Aug 1 05:57:30 user.debug kernel: PCI(00:08.0): version 0.8
* Aug 1 05:57:30 user.debug kernel: PCI: Setting latency timer of device 00:08.0 to 64
* Aug 1 05:57:30 user.debug kernel: PCI: Setting latency timer of device 00:08.1 to 64
* Aug 1 05:57:30 user.info kernel: ata1: SATA max UDMA/133 cmd 0xD480 ctl 0xD402 bmdma 0xCC00 irq 15
* Aug 1 05:57:30 user.info kernel: ata2: SATA max UDMA/133 cmd 0xD080 ctl 0xD002 bmdma 0xCC08 irq 15
* Aug 1 05:57:30 user.debug kernel: ata1: dev 0 cfg 49:2f00 82:746b 83:7f09 84:4123 85:7469 86:bc01 87:4123 88:40ff
* Aug 1 05:57:30 user.info kernel: ata1: dev 0 ATA-8, max UDMA7, 312581808 sectors: LBA48
* Aug 1 05:57:30 user.info kernel: ata1: dev 0 configured for UDMA/133
* Aug 1 05:57:30 user.debug kernel: ata2: dev 0 cfg 49:2f00 82:746b 83:7f09 84:4123 85:7469 86:bc01 87:4123 88:40ff
* Aug 1 05:57:30 user.info kernel: ata2: dev 0 ATA-8, max UDMA7, 312581808 sectors: LBA48
* Aug 1 05:57:30 user.info kernel: ata2: dev 0 configured for UDMA/133
* Aug 1 05:57:30 user.info kernel: ata3: SATA max UDMA/133 cmd 0xC880 ctl 0xC802 bmdma 0xC080 irq 5
* Aug 1 05:57:30 user.info kernel: ata4: SATA max UDMA/133 cmd 0xC480 ctl 0xC402 bmdma 0xC088 irq 5
* Aug 1 05:57:30 user.debug kernel: ata3: dev 0 cfg 49:2f00 82:746b 83:7f09 84:4123 85:7469 86:bc01 87:4123 88:40ff
* Aug 1 05:57:30 user.info kernel: max UDMA7, 312581808 sectors: LBA48
* Aug 1 05:57:30 user.info kernel: ata3: dev 0 configured for UDMA/133
* Aug 1 05:57:30 user.info kernel: ata4: no device found (phy stat 00000000)
* Aug 1 05:57:30 user.info kernel: scsi1 : sata_nv
* Aug 1 05:57:30 user.info kernel: scsi2 : sata_nv
* Aug 1 05:57:30 user.info kernel: scsi3 : sata_nv
* Aug 1 05:57:30 user.info kernel: scsi4 : sata_nv
* Aug 1 05:57:30 user.warn kernel: Vendor: ATA Model: SAMSUNG HD161HJ Rev: JF10
* Aug 1 05:57:30 user.warn kernel: Type: Direct-Access ANSI SCSI revision: 05
* Aug 1 05:57:30 user.warn kernel: Vendor: ATA Model: SAMSUNG HD161HJ Rev: JF10
* Aug 1 05:57:30 user.warn kernel: Type: Direct-Access ANSI SCSI revision: 05
* Aug 1 05:57:30 user.warn kernel: Vendor: ATA Model: SAMSUNG HD161HJ Rev: JF10
* Aug 1 05:57:30 user.warn kernel: Type: Direct-Access ANSI SCSI revision: 05
* Aug 1 05:57:30 user.warn kernel: Attached scsi disk sda at scsi1, channel 0, id 0, lun 0
* Aug 1 05:57:30 user.warn kernel: Attached scsi disk sdb at scsi2, channel 0, id 0, lun 0
* Aug 1 05:57:30 user.warn kernel: Attached scsi disk sdc at scsi3, channel 0, id 0, lun 0
* Aug 1 05:57:30 user.warn kernel: SCSI device sda: 312581808 512-byte hdwr sectors (160042 MB)
* Aug 1 05:57:30 user.info kernel: Partition check:
* Aug 1 05:57:30 user.info kernel: /dev/scsi/host1/bus0/target0/lun0: p1
* Aug 1 05:57:30 user.warn kernel: SCSI device sdb: 312581808 512-byte hdwr sectors (160042 MB)
* Aug 1 05:57:30 user.info kernel: /dev/scsi/host2/bus0/target0/lun0: p1
* Aug 1 05:57:30 user.warn kernel: SCSI device sdc: 312581808 512-byte hdwr sectors (160042 MB)
* Aug 1 05:57:30 user.info kernel: /dev/scsi/host3/bus0/target0/lun0: p1
* Aug 1 05:57:30 user.info kernel: raw1394: /dev/raw1394 device initialized
* Aug 1 05:57:30 user.info kernel: sbp2: $Rev: 1074 $ Ben Collins
* Aug 1 05:57:30 user.err kernel: ieee1394: sbp2: Driver forced to serialize I/O (serialize_io = 1)
* Aug 1 05:57:30 user.info kernel: usb.c: registered new driver usbdevfs
* Aug 1 05:57:30 user.info kernel: usb.c: registered new driver hub
* Aug 1 05:57:30 user.info kernel: host/uhci.c: USB Universal Host Controller Interface driver v1.1
* Aug 1 05:57:30 user.debug kernel: PCI: Setting latency timer of device 00:02.0 to 64
* Aug 1 05:57:30 user.info kernel: host/usb-ohci.c: USB OHCI at membase 0xf881a000, IRQ 10
* Aug 1 05:57:30 user.info kernel: host/usb-ohci.c: usb-00:02.0, PCI device 10de:03f1
* Aug 1 05:57:30 user.info kernel: usb.c: new USB bus registered, assigned bus number 1
* Aug 1 05:57:30 user.info kernel: hub.c: USB hub found
* Aug 1 05:57:30 user.info kernel: hub.c: 10 ports detected
* Aug 1 05:57:30 user.info kernel: usb.c: registered new driver usbkbd
* Aug 1 05:57:30 user.info kernel: usbkbd.c: :USB HID Boot Protocol keyboard driver
* Aug 1 05:57:30 user.info kernel: Initializing USB Mass Storage driver.
* Aug 1 05:57:30 user.info kernel: usb.c: registered new driver usb-storage
* Aug 1 05:57:30 user.info kernel: USB Mass Storage support registered.
* Aug 1 05:57:30 user.info kernel: NET4: Linux TCP/IP 1.0 for NET4.0
* Aug 1 05:57:30 user.info kernel: IP Protocols: ICMP, UDP, TCP
* Aug 1 05:57:30 user.info kernel: IP: routing cache hash table of 8192 buckets, 64Kbytes
* Aug 1 05:57:30 user.info kernel: TCP: Hash tables configured (established 262144 bind 65536)
* Aug 1 05:57:30 user.info kernel: NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
* Aug 1 05:57:30 user.notice kernel: RAMDISK: NASLite file system found at block 0
* Aug 1 05:57:30 user.warn kernel: <6>hub.c: new USB device 00:02.0-4, assigned address 2
* Aug 1 05:57:30 user.warn kernel: .<6>scsi5 : SCSI emulation for USB Mass Storage devices
* Aug 1 05:57:30 user.warn kernel: . Vendor: LEXAR Model: JUMPDRIVE ELITE Rev: 1000
* Aug 1 05:57:30 user.warn kernel: Type: Direct-Access ANSI SCSI revision: 02
* Aug 1 05:57:30 user.warn kernel: Attached scsi removable disk sdd at scsi5, channel 0, id 0, lun 0
* Aug 1 05:57:30 user.warn kernel: <6>Freeing initrd memory: 1621k freed
* Aug 1 05:57:30 user.warn kernel: VFS: Mounted root (ext2 filesystem).
* Aug 1 05:57:30 user.info kernel: Mounted devfs on /dev
* Aug 1 05:57:30 user.info kernel: Freeing unused kernel memory: 536k freed
* Aug 1 05:57:30 user.warn kernel: SCSI device sdd: 506880 512-byte hdwr sectors (260 MB)
* Aug 1 05:57:30 user.warn kernel: sdd: Write Protect is off
* Aug 1 05:57:30 user.info kernel: /dev/scsi/host5/bus0/target0/lun0: p1
* Aug 1 05:57:30 user.debug kernel: WARNING: USB Mass Storage data integrity not assured
* Aug 1 05:57:30 user.debug kernel: USB Mass Storage device found at 2
* Aug 1 05:57:30 user.info kernel: kjournald starting. Commit interval 5 seconds
* Aug 1 05:57:30 user.info kernel: EXT3 FS 2.4-0.9.19, 19 August 2002 on sd(8,1), internal journal
* Aug 1 05:57:30 user.info kernel: EXT3-fs: mounted filesystem with ordered data mode.
* Aug 1 05:57:30 user.info kernel: kjournald starting. Commit interval 5 seconds
* Aug 1 05:57:30 user.info kernel: EXT3 FS 2.4-0.9.19, 19 August 2002 on sd(8,17), internal journal
* Aug 1 05:57:30 user.info kernel: EXT3-fs: mounted filesystem with ordered data mode.
* Aug 1 05:57:30 user.info kernel: kjournald starting. Commit interval 5 seconds
* Aug 1 05:57:30 user.info kernel: EXT3 FS 2.4-0.9.19, 19 August 2002 on sd(8,33), internal journal
* Aug 1 05:57:30 user.info kernel: EXT3-fs: mounted filesystem with ordered data mode.

How are you benchmarking this?

Did you try this with a crossover cable as i suggested in email to eliminate the network from the problem?

Turn off ACPI.

Hi Ralph.

Thanks for the fast answer. I don't have a cross cable under the hand so couldn't test it yet but I don't think it to be LAN/cable/switch related as all work fast and constant for all the reads and reads and writes under FTP and XP to XP.

The "Benchmarking" is done by an "intelligent" copy program SuperCopier2 which enhences the "stupid" copy of XP. I have done the tests shutting down SuperCopier2 and using NetMeter and the writes to the NL server have always "hearthbeat" style speeds (spike fast, slow, slow, slow, spike fast...) but the reads are all constantly high.

I have stopped ACPI in the BIOS but no change.

This mornings test (all clear boot, even the router and switch): First copy from computer 3 to NL: 30000 kB/s, second copy right after the first one, 12000 kB/s . Third copy 12000kB/s again. Copy from Computer 2 to NL 1000 - 500 kB/s (and comp 2 is a fast computer with dual channer RAM, raid 0 etc). Almost never goes much higher. Several copies from computer 1 to NL, all around 12000 kB/s. Read from NL to all computers between 28 and 38 000 kB/s.

I have to test LAN settings of computer 2 to see why it is constantly slower than the other 2 but I still don't find the reason of the the write speed changes.

I have switched the position of the cables on the swich and nothing changed. For the image below, on computer 2, I have shut down SuperCopier and mesured with NetMeter (image below). The highest spike is 9 MiB/s, the other spikes are around 3,5 MiB/s but the main part is less than 1 MiB/s



This second image (below) is done with computer 1. The first red is the read of a 250MB file (the same test file on all computers) from the NL server and is on the 39 MiB/s scale But the green part is the write of the same file from computer 1 to the NL server. As you can see, the difference is huge (it is on the 3-4 MiB/s avarage).



This ends this mornings tests. If you want me to test something else, please go ahead.

.../... Follow-up: I have decided to try to use NL server for my dental software to test it for real. Well, speed sucks Everything started to slow down VERY significantly (lvalidation of one treatement took less than a sec with XP Pro as a server and with NL, it takes +/- 5 seconds). We have used NL as a server for 3 hours and I had to change back to XP Pro as it was unusable