Link to home
Start Free TrialLog in
Avatar of ourpassions
ourpassions

asked on

Partial internet connectivity under linux while win98 gives full connectivity on same machine

i access internet using lan. i have windows 98 and linux red hat 7.3 installed. While i can surf any website and use yahoo messenger in windows 98, i cannot surf most of the sites in linux and cannot use yahoo messenger. Only some sites like google.com can be accessed by both linux and windows.I can get my mails using email client in both linux and windows properly.

On analying TCP Dumps , i figured out that my machine sends TCP Reset often..in linux while in windows it sends it rarely . In general internet connection under  window is good while under linux is pathetic!!!

What to do so that i get some internet connectivity in linux as i get in windows?

highly in need of your idea, suggestions
-vishal
Avatar of jlevie
jlevie

What ethernet interface (make/model) do you have, what version of Linux, and how do you connect to the Internet (cable, adsl modem, etc)?

It sounds like there's a basic networking problem when the system is running Linux. That might be a result of a driver problem, some incompatibility between the ethernet interface and your Internet modem/router, or something about the BIOS/motherboard that Linux doesn't like.
The above information is important to help further.  Please note that a lot of pain has been felt using some 10 mb ethernet cards plugged into some Linksys hubs (particularly on the back of DSL routers).  I've felt this pain personally as well.  The link status would bounce all the time making any connectivity with this particular system a nightmare.  Installing a 100mb card fixed the problem.  I believed this to be a Linksys/10mb nic problem, but having never tested it in Windows I can't say for sure.
Avatar of ourpassions

ASKER

I have  D-Link DFE-538TX 10/100 Adapter and i use lan to connect internet. The router at other end is DI 1662 Dlink router. They have also put guardianpro firewall before router. I have Redhat Linux 7.3.
If it was a driver problem, then i guess i would not be getting even parital connectivity. i can ping www.yahoo.com but cannot connect to it using browser. where as i can ping and connect to www.google.com site. So i do have connectivity but partial? Ok Even if its a driver incompatibility problem , how do i solve it ? Specific steps will be highy appreciated.

I have caputured tcpdumps in linux and windows and saw that under linux there are lot of TCP RST where as rare in win 98. To download these tcpdumps pls visit http://groups.yahoo.com/group/BNPGs/files/   and dont forget to read the "readme_first.html"

-vishal
7.3 & the D-Link NIC certainly should be okay if the 7.3 system is up to date w/respect to the RedHat errata. And the kernel from the CD may be okay.  However, there still could be what's in essence a driver problem if the direver/NIC is mis-negotiating the link speed/mode. If this system connects to a 100Mbps hub or switch it may be in the wrong mode. Such a link will appear to work, if a bit slowly, but will exhibit errors on larger data transfers. You can get a diagnostic tool from http://www.scyld.com/diag that can be used to display the link characteristics. I think you are using the ne driver for that card (lsmod will tell).

If you connect to a hub the link should be 100/HDX and a switch should probably be 100/FDX.

It is also possible that this is an MTU problem. To test that you could use ifconfg to set the MTU to something like 1300.
I have tried all the suggestions like changing NIC , changing MTU whatever was mentioned by everybody here. i even reinstalled Linux  RH 7.3 but the problem still exist.
I have lost hope! But i have now increased points to 500 for this problem
: (
What NIC do you have?

Did you try running the diagnostic to see if the negotiated link parameters are correct?
can you please post the output of "dmesg" (must be executed from a command line).
You can capture this by executing like this:
"# dmesg > output.txt"
Then, copy and paste the file or upload it.
Also, please scan through the end of the file /var/log/messages in your favorite editor (only as root).  Look for problems.  You are describing behavior I have never heard of so it sounds like a misconfiguration or corrupted something.  RESET's are not typically sent out for no reason.  Could you include a summary of the sniff (or attach a short one) so we can see it?  Seeing RESET's can be a normal event, given the right circumstances.  Windows often wouldn't send resets in the right times because their IP Stack is buggy.

Also, please include the output of "ifconfig -a" after  you have been running for a little bit and trying to surf the web.  Also, if you can try FTP it'd be great.

Are you running the RedHat included Firewall?  If so, please disable it and see if your problem goes away.

Thanks,
Matt
I have DFE-538TX from Dlink. I running the diagnostic (rtl8139-diag)to see if the negotiated link parameters are correct. But how do i know whether link parameters are correct? i

OUTPUT OF rtl8139-diag.o with -a
=================================
rtl8139-diag.c:v2.11 4/22/2003 Donald Becker (becker@scyld.com)
 http://www.scyld.com/diag/index.html
Index #1: Found a D-Link DFE-530TX+ (RealTek RTL8139C) adapter at 0xd000.
The RealTek chip appears to be active, so some registers will not be read.
To see all register values use the '-f' flag.
RealTek chip registers at 0xd000
 0x000: 425d0500 00007651 80000000 00000000 9008a078 9008a078 9008a03c 9008a04a
 0x020: 0c924000 0c924600 0c924c00 0c925200 0c910000 0d0a0000 22f822e8 0000c07f
 0x040: 74000680 0000f78e d3aa423b 00000000 008d1000 00000000 0088c118 00100000
 0x060: 1000f00f 01e1782d 00000000 00000000 00000005 000f77c0 b0f243b9 7a36d743.
Realtek station address 00:05:5d:42:51:76, chip type 'rtl8139C'.
  Receiver configuration: Normal unicast and hashed multicast
     Rx FIFO threshold 2048 bytes, maximum burst 2048 bytes, 32KB ring
  Transmitter enabled with NONSTANDARD! settings, maximum burst 1024 bytes.
    Tx entry #0 status 9008a078 complete, 120 bytes.
   Tx carrier lost
    Tx entry #1 status 9008a078 complete, 120 bytes.
   Tx carrier lost
    Tx entry #2 status 9008a03c complete, 60 bytes.
   Tx carrier lost
    Tx entry #3 status 9008a04a complete, 74 bytes.
   Tx carrier lost
  Flow control: Tx disabled  Rx disabled.
  The chip configuration is 0x10 0x8d, MII half-duplex mode.
  No interrupt sources are pending.
rtl8139-diag.c:v2.11 4/22/2003 Donald Becker (becker@scyld.com)
 http://www.scyld.com/diag/index.html


OUTPUT OF rtl8139-diag.o with -m
=================================
Index #1: Found a D-Link DFE-530TX+ (RealTek RTL8139C) adapter at 0xd000.
Realtek station address 00:05:5d:42:51:76, chip type 'rtl8139C'.
  Receiver configuration: Normal unicast and hashed multicast
     Rx FIFO threshold 2048 bytes, maximum burst 2048 bytes, 32KB ring
  Transmitter enabled with NONSTANDARD! settings, maximum burst 1024 bytes.
  Flow control: Tx disabled  Rx disabled.
  The chip configuration is 0x10 0x8d, MII half-duplex mode.
  No interrupt sources are pending.
 The RTL8139 does not use a MII transceiver.
 It does have internal MII-compatible registers:
   Basic mode control register   0x1000.
   Basic mode status register    0x782d.
   Autonegotiation Advertisement 0x01e1.
   Link Partner Ability register 0x0000.
   Autonegotiation expansion     0x0000.
   Disconnects                   0x0000.
   False carrier sense counter   0x0000.
   NWay test register            0x0005.
   Receive frame error count     0x0000.



OUTPUT OF rtl8139-diag.o with -e
=================================
rtl8139-diag.c:v2.11 4/22/2003 Donald Becker (becker@scyld.com)
 http://www.scyld.com/diag/index.html
Index #1: Found a D-Link DFE-530TX+ (RealTek RTL8139C) adapter at 0xd000.
Realtek station address 00:05:5d:42:51:76, chip type 'rtl8139C'.
  Receiver configuration: Normal unicast and hashed multicast
     Rx FIFO threshold 2048 bytes, maximum burst 2048 bytes, 32KB ring
  Transmitter enabled with NONSTANDARD! settings, maximum burst 1024 bytes.
  Flow control: Tx disabled  Rx disabled.
  The chip configuration is 0x10 0x8d, MII half-duplex mode.
  No interrupt sources are pending.
Decoded EEPROM contents:
   PCI IDs -- Vendor 0x1186, Device 0x1300.
   PCI Subsystem IDs -- Vendor 0x1186, Device 0x1300.
   PCI timer settings -- minimum grant 32, maximum latency 64.
  General purpose pins --  direction 0xe1  value 0x12.
  Station Address 00:05:5D:42:51:76.
  Configuration register 0/1 -- 0x8d / 0xc2.
 EEPROM active region checksum is 0712.
i do not have firewall running at my linux . i did sniff but it gives no advantage or better readability . i prefer etheral . I have already uploaded my tcpdumps  at http://groups.yahoo.com/group/BNPGs/files/   

This message contails
1) OUTPUT OF dmesg
2) OUTPUT OF ifconfig -a

1) OUTPUT OF dmesg
===================
Linux version 2.4.18-3 (bhcompile@porky.devel.redhat.com) (gcc version 2.96 20000731 (Red Hat Linux 7.3 2.96-110)) #1 Thu Apr 18 07:32:41 EDT 2002
BIOS-provided physical RAM map:
 BIOS-e820: 0000000000000000 - 00000000000a0000 (usable)
 BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved)
 BIOS-e820: 0000000000100000 - 000000000fff0000 (usable)
 BIOS-e820: 000000000fff0000 - 000000000fff3000 (ACPI NVS)
 BIOS-e820: 000000000fff3000 - 0000000010000000 (ACPI data)
 BIOS-e820: 00000000ffff0000 - 0000000100000000 (reserved)
On node 0 totalpages: 65520
zone(0): 4096 pages.
zone(1): 61424 pages.
zone(2): 0 pages.
Kernel command line: ro root=/dev/hda3
Local APIC disabled by BIOS -- reenabling.
Found and enabled local APIC!
Initializing CPU#0
Detected 1330.250 MHz processor.
Console: colour VGA+ 80x25
Calibrating delay loop... 2654.20 BogoMIPS
Memory: 255356k/262080k available (1125k kernel code, 6340k reserved, 802k data, 296k init, 0k highmem)
Dentry cache hash table entries: 32768 (order: 6, 262144 bytes)
Inode cache hash table entries: 16384 (order: 5, 131072 bytes)
Mount-cache hash table entries: 4096 (order: 3, 32768 bytes)
Buffer cache hash table entries: 16384 (order: 4, 65536 bytes)
Page-cache hash table entries: 65536 (order: 6, 262144 bytes)
CPU: Before vendor init, caps: 0383fbff c1c3fbff 00000000, vendor = 2
CPU: L1 I Cache: 64K (64 bytes/line), D cache 64K (64 bytes/line)
CPU: L2 Cache: 256K (64 bytes/line)
CPU: After vendor init, caps: 0383fbff c1c3fbff 00000000 00000000
Intel machine check architecture supported.
Intel machine check reporting enabled on CPU#0.
CPU:     After generic, caps: 0383fbff c1c3fbff 00000000 00000000
CPU:             Common caps: 0383fbff c1c3fbff 00000000 00000000
CPU: AMD Athlon(tm) XP 1500+ stepping 02
Enabling fast FPU save and restore... done.
Enabling unmasked SIMD FPU exception support... done.
Checking 'hlt' instruction... OK.
POSIX conformance testing by UNIFIX
mtrr: v1.40 (20010327) Richard Gooch (rgooch@atnf.csiro.au)
mtrr: detected mtrr type: Intel
PCI: PCI BIOS revision 2.10 entry at 0xfb460, last bus=1
PCI: Using configuration type 1
PCI: Probing PCI hardware
Unknown bridge resource 0: assuming transparent
PCI: Using IRQ router default [1106/3099] at 00:00.0
isapnp: Scanning for PnP cards...
isapnp: No Plug & Play device found
Linux NET4.0 for Linux 2.4
Based upon Swansea University Computer Society NET3.039
Initializing RT netlink socket
apm: BIOS version 1.2 Flags 0x07 (Driver version 1.16)
Starting kswapd
VFS: Diskquotas version dquot_6.5.0 initialized
Detected PS/2 Mouse Port.
pty: 2048 Unix98 ptys configured
Serial driver version 5.05c (2001-07-08) with MANY_PORTS MULTIPORT SHARE_IRQ SERIAL_PCI ISAPNP enabled
ttyS00 at 0x03f8 (irq = 4) is a 16550A
ttyS01 at 0x02f8 (irq = 3) is a 16550A
Real Time Clock Driver v1.10e
block: 496 slots per queue, batch=124
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
VP_IDE: IDE controller on PCI bus 00 dev 89
PCI: No IRQ known for interrupt pin A of device 00:11.1. Please try using pci=biosirq.
VP_IDE: chipset revision 6
VP_IDE: not 100% native mode: will probe irqs later
VP_IDE: Unknown VIA SouthBridge, contact Vojtech Pavlik <vojtech@suse.cz>
hda: SAMSUNG SV4002H, ATA DISK drive
hdc: SAMSUNG CD-ROM SC-152C, ATAPI CD/DVD-ROM drive
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide0: probed IRQ 14 failed, using default.
ide1 at 0x170-0x177,0x376 on irq 15
ide1: probed IRQ 15 failed, using default.
hda: 78242976 sectors (40060 MB) w/1945KiB Cache, CHS=4870/255/63
ide-floppy driver 0.99.newide
Partition check:
 hda: hda1 hda2 < hda5 hda6 > hda3 hda4
Floppy drive(s): fd0 is 1.44M
FDC 0 is a post-1991 82077
RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize
ide-floppy driver 0.99.newide
md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
md: Autodetecting RAID arrays.
md: autorun ...
md: ... autorun DONE.
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP, IGMP
IP: routing cache hash table of 2048 buckets, 16Kbytes
TCP: Hash tables configured (established 16384 bind 16384)
Linux IP multicast router 0.06 plus PIM-SM
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
RAMDISK: Compressed image found at block 0
Freeing initrd memory: 125k freed
VFS: Mounted root (ext2 filesystem).
Journalled Block Device driver loaded
kjournald starting.  Commit interval 5 seconds
EXT3-fs: mounted filesystem with ordered data mode.
Freeing unused kernel memory: 296k freed
Adding Swap: 594396k swap-space (priority -1)
usb.c: registered new driver usbdevfs
usb.c: registered new driver hub
usb-uhci.c: $Revision: 1.275 $ time 07:48:41 Apr 18 2002
usb-uhci.c: High bandwidth mode enabled
usb-uhci.c: USB UHCI at I/O 0xd800, IRQ 9
usb-uhci.c: Detected 2 ports
usb.c: new USB bus registered, assigned bus number 1
hub.c: USB hub found
hub.c: 2 ports detected
usb-uhci.c: USB UHCI at I/O 0xdc00, IRQ 9
usb-uhci.c: Detected 2 ports
usb.c: new USB bus registered, assigned bus number 2
hub.c: USB hub found
hub.c: 2 ports detected
usb-uhci.c: v1.275:USB Universal Host Controller Interface driver
EXT3 FS 2.4-0.9.17, 10 Jan 2002 on ide0(3,3), internal journal
ide-floppy driver 0.99.newide
hdc: ATAPI 52X CD-ROM drive, 128kB Cache
Uniform CD-ROM driver Revision: 3.12
parport0: PC-style at 0x378 [PCSPP,TRISTATE]
8139too Fast Ethernet driver 0.9.24
eth0: D-Link DFE-538TX (RealTek RTL8139) at 0xd08bc000, 00:05:5d:42:51:76, IRQ 12
eth0:  Identified 8139 chip type 'RTL-8139C'
eth0: Setting half-duplex based on auto-negotiated partner ability 0000.
Via 686a audio driver 1.9.1
ac97_codec: AC97 Audio codec, id: 0x414c:0x4730 (Unknown)
via82cxxx: Codec rate locked at 48Khz
via82cxxx: board #1 at 0xE000, IRQ 5
cdrom: This disc doesn't have any tracks I recognize!


2) OUTPUT OF ifconfig -a
=====================
eth0      Link encap:Ethernet  HWaddr 00:05:5D:42:51:76  
          inet addr:90.0.0.232  Bcast:90.0.0.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:157237 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2082 errors:0 dropped:0 overruns:0 carrier:0
          collisions:39 txqueuelen:100
          RX bytes:11736538 (11.1 Mb)  TX bytes:251007 (245.1 Kb)
          Interrupt:12 Base address:0xc000

lo        Link encap:Local Loopback  
          inet addr:127.0.0.1  Mask:255.0.0.0
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:50 errors:0 dropped:0 overruns:0 frame:0
          TX packets:50 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:3288 (3.2 Kb)  TX bytes:3288 (3.2 Kb)
Thanks for the information.  Sorry to make you tell me about the sniff location again.  I'm sure you are probably quite frustrated at the moment, but don't worry, we'll figure this out.

This is quite interesting:
  Transmitter enabled with NONSTANDARD! settings, maximum burst 1024 bytes.
    Tx entry #0 status 9008a078 complete, 120 bytes.
   Tx carrier lost
    Tx entry #1 status 9008a078 complete, 120 bytes.
   Tx carrier lost
    Tx entry #2 status 9008a03c complete, 60 bytes.
   Tx carrier lost
    Tx entry #3 status 9008a04a complete, 74 bytes.
   Tx carrier lost

Although on its own it doesn't explain much.  How is this system connected to your ISP?  eth0 plugs directly into what?  Did you happen to unplug and replug in your network cable during this system's uptime that it would show this?
I'm interested in the "NONSTANDARD settings" message as well.  Could this be something you have changed or something automagically set for you by RH?

Looking at the sniffs showed me the RST packets immediately following the SYN/ACK from the server.  The server itself was a different server on a different network between the sniffs.  If would be helpful to see sniffs for Yahoo, actually picking which IP address you visit.  You can get a list by typing the following command from the command line:
# host www.yahoo.com
Pick one of the servers listed (216.109.118.70 for example) and type that address in for both Win and Linux.  The reason I'm asking for this is that the signature is different (The SYN/ACK packets from the server are very different) and while I'm not saying you aren't having a problem, I would like to have apples to compare...  It's difficult to see subtle differences when there are many not so subtle ones caused by different OS versions on the servers.

The sniff itself doesn't tell much, since there are no long time-outs or anything that would ordinarily cause a Reset.  It suggests that there is still something in the OS which is causing these.  This is a Layer-4 problem at this point.  Just for my peace-of-mind, please send the output from:

# iptables -L -n
     and
# iptables -t nat -L -n
     and
# iptables -t mangle -L -n

This will dump the NetFilter information, just in case there is something that sticks in filters without tell you.  The most common reasons you'll see a RESET are:
* System is overloaded and can't handle another connection
* Parameters are not acceptible when setting up a TCP connection (this is why the sniffs need to be to the same server, because the server's values are different between the LIN and WIN sniffs)
* Tampering or failure in the network which causes something to not be right (like a Sequence number being wrong)
* Packets arriving after the connection has already been shut down
* NetFilter's REJECT target has been reached (NetFilter is the internal Linux firewall subsystem, available and often used on every version of Linux since the 2.4 kernel)

Sorry we don't have answers yet, but the correspondance method is not very speed when it comes to evaluating a LOT of information like we are doing right now.  If only I could SSH to the system we could look at a lot of stuff quickly.

Speaking of SSH, are you able to do other things, like SSH, telnet and ftp?  You said that email worked but IM and WWW don't.  Do you have other PC's on the local network that you can test communication to and eliminate the ISP's firewall and routers from the equation?

Thank you for your patience.  I hate the waiting as well.  I would like to find it as badly as you would like it found :)
My machine is connected to a Dlink switch . There are a number of such switches connecting to the router which in turn connect to a server running GuardianPro Firewall.

i didnt plug or unplug while running the rtl8139-diag.o with -a. It shows me all the time NONSTANDARD setting and TX lost.
I have uploaded following additional files at http://groups.yahoo.com/group/BNPGs/files/ 

1) iptables.out : contains output of iptables command run under my linux
2) mii-diag.out : contains ourput of mii-diag program which says that ***link status not establised***
3) yahoo_win : packet capture when i try to connect to 216.109.118.70  from windows 98
4) yahoo_lin : packet capture when i try to connect to 216.109.118.70  from linux RH 7.3
Pls note that my WWW works under linux but only partially. I can visit webmail.vsnl.com or www.google.coo or www.mit.edu etc. where as i cannot access www.yahoo.com or www.rediffmail.com or .....(around >80% i cannot visit).

Also if i use dial up then i can access any website and everything works fine. It is only when i use cable net using my ethernet card and use linux doesn this problem arise.
I have asked my cable guy who provides me cable internet to give me a live ip address for some time so that any one can ssh to me and debug the problem in my machine.he has said that the technical guy will come after 1 week and then he will do it for me.

Also thanks for motivating response..i had nearly lost all the hopes to solve this problem.
-vishal
What D-Link switch do you have? According to rtl8139-diag your link has been forced to 10Mbps/HDX (Auto-negotiation disabled), which would be correct if you are connecting to a 10Mbps only hub or switch. While that should also work if you connect to a 10/100 switch it is possible that there could be problems in such a configuration.

Did you use the diag tool to force the link the 10Mbps/HDX? Have you tried it with autonegotiation enabled?
hi jlevie,
I am a newbie in linux . so more detailed suggestions from you would be really helpful.
1) How do i check whether i m connected to 10Mbps or 100 Mbps switch?
2) How do i find whether there could be problem in such a configuration?

I didnt force the link to be 10Mbps. I did try to enable autonegotiation but it didnt happened.

-vishal
This linux 7.3 is now is frustrating me a lot...Help!
(1) If it doesn't say so on the switch label, get the make/model info and look it up on the web. Or tell me and I'll research it. I tried a search for a "DI 1662" (referenced earlier as the router) but failed to find anything.

I think the command would be './rtl8139-diag -f -A 100baseTx' to enable autonegotiation.

I'm sorry you are having problems, but I really don't think that it is a flaw in 7.3, but rather more related to the ethernet interface and your network. Personally I prefer the higher quality NIC's, like the Intel EtherExpress or 3Com cards, mainly because I find they have less interoperability issues. At the lower end I've seen pretty good results with the Netgear cards.

i did ./rtl8139-diag -f -A 100baseTx but it didnt enable autonegotiation. Now only thing left is let my cable internet guy give me a live ip address and then experts here  can remotely ssh to my machine and find out the problem.
I will get live ip address after 1 week.
If any one is interest in doing this for me.. mail me your email address

-vishal
Since that didn't cause the card to advertise auto negotiation I'm wondering if there's something wrong with that card. Would it be possible to swap the card for a different brand, preferrably an Intel or 3Com?
it is not possible to swap with other card as i dont have money to buy new one. But i will try if i can temporarily get some Intel or 3Com card. I m confused  now.

If my card is faulty then why does it work properly under win 98 giving me excellent speed? More than the card, it could be driver problem?

-vishal
Hmmm....  I'm not sure whether I'm supposed to give this, but matt_AT_eisgr.com.
Have you tried another browser?  Like Konqueror?  This is still a wierd issue that looks like it is anywhere from Layer 4+ and the browser could be part of it...  I'm still interested to see if you can do many other things.  If you can do Email ok, then I would believe that the problem is indeed with the browser or one of its libraries (if the problem is with your machine at all... It still sounds like the ISP's network is a question).  What name and version browser are you using?
The reason that I suggested swapping the NIC is to see if the problem goes away with a different NIC. If it does that might mean that the RH 7.3 driver for that particular card doesn't work correctly. D-Link network cards seem to frequently interact poorly with Linux. As I recall there've been a number of questions on EE about difficulties with the D-Link NIC's.

If it turns out to be related to the D-Link card, updating the system with the 7.3 errata (as in running up2date) might help as you'll get a much later kernel and drivers.
i have tried in different browsers . same thing happens SYN,SYN ACK,RST...
also i will try to up2date in linux..

-vishal
Good thought.  I might agree with jlevie, except for the TCP RESET's.  I may be incorrect in this, but that typically doesn't happen consistently in some communication when the problem is at layer 2 or 1.  Ordinarily Layer 1 and 2 problems cause the sniffs to look very sporatic and the problems would be seen across the board.

Do you have a friend with a laptop and Linux who can bring their machine over and test your Internet connection?  Conversely, can you take your PC to someone else's broadband connection (with a DIFFERENT ISP) and test against that?  

Thanks for your patience.  Let me know if you would like to attempt a remote session.
Special Thanks jlevie  and r4780y for the both of your suggestions. Although we couldnt reach to the root of the problem or even get a working solution, i learned a lot about linux networking from both of you.

My  senior in my company (I work for a network security firm as trainee) finally solved my problem. He suggested me to just change the tcp_sack and tcp_timestamps paratmeter to 0. and Wow!! Now everthing is working fine.

Thanks a lot
vishal gupta
ourpassions@yahoo.com
mrvishalgupta@hotmail.com
No problem.  Glad to hear it is working.  Point of interest...  Why was this necessary?  My systems included TCP_SACK and TCP_TIMESTAMPS as well and everything worked...  It still sounds like your ISP's firewall was doing something odd.  Either way, good luck!
I agree with r4780y, changing those settings should not be necessary for any reasonably sane firewall. I've worked with Linux boxes behind a lot of different firewalls and I've never seen a problem that could be solved like this. What firewall is between yo and the Internet?
guardianpro firewall is between my machine and internet. I found that TCP_SACK is not a problem at all. only TCP_TIMESTAMPS is a problem. so i turned it off for time being. i will find out the root case and announce here in this message thread to conclude this disscussion.
Thanks guys once again
-vishal
ASKER CERTIFIED SOLUTION
Avatar of GhostMod
GhostMod
Flag of United States of America image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial