Problem. [Archive] - SpeedGuide.net Broadband Community

View Full Version : Problem.


ps2wiz
07-10-03, 10:55 PM
Ok, I have noticed a problem with my tweaks. When I run the tcp-ip analzer in Internet Explorer i get these results.

TCP options string = 020405b401010402
MTU = 1500
MTU is fully optimized for broadband.
MSS = 1460
Maximum useful data in each packet = 1460, which equals MSS.
Default Receive Window (RWIN) = 8760
RWIN Scaling (RFC1323) = 0 bits
Unscaled Receive Window = 8760

RWIN is a multiple of MSS
Other values for RWIN that might work well with your current MTU/MSS:
513920 (MSS x 44 * scale factor of 8)
256960 (MSS x 44 * scale factor of 4)
128480 (MSS x 44 * scale factor of 2)
64240 (MSS x 44)
bandwidth * delay product:

Your RcvWindow limits you to: 350.4 kbps (43.8 KBytes/s) @ 200ms
Your RcvWindow limits you to: 140.16 kbps (17.52 KBytes/s) @ 500ms
Consider increasing your RWIN value to optimize TCP/IP for broadband.
MTU Discovery (RFC1191) = ON
Time to live left = 121 hops

TTL value is ok.
Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON
IP type of service field (RFC1349) = 00000000

____________________________________________________

And a second later the page itself automatically refreshes and shows these results, which my intended results after the tweaks.

____________________________________________________
TCP options string = 020405b401010402
MTU = 1500
MTU is fully optimized for broadband.
MSS = 1460
Maximum useful data in each packet = 1460, which equals MSS.
Default Receive Window (RWIN) = 8760
RWIN Scaling (RFC1323) = 0 bits
Unscaled Receive Window = 8760

RWIN is a multiple of MSS
Other values for RWIN that might work well with your current MTU/MSS:
513920 (MSS x 44 * scale factor of 8)
256960 (MSS x 44 * scale factor of 4)
128480 (MSS x 44 * scale factor of 2)
64240 (MSS x 44)
bandwidth * delay product:

Your RcvWindow limits you to: 350.4 kbps (43.8 KBytes/s) @ 200ms
Your RcvWindow limits you to: 140.16 kbps (17.52 KBytes/s) @ 500ms
Consider increasing your RWIN value to optimize TCP/IP for broadband.
MTU Discovery (RFC1191) = ON
Time to live left = 121 hops

TTL value is ok.
Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON
IP type of service field (RFC1349) = 00000000

____________________________________________________

But i dont use Internet Explorer Because I have found mozilla to be faster, and when I run the tcp-ip test I get the same results as the the one in Internet Explorer without it refreshes, in case you are too lazy to scroll up here they are again.
____________________________________________________

TCP options string = 020405b401010402
MTU = 1500
MTU is fully optimized for broadband.
MSS = 1460
Maximum useful data in each packet = 1460, which equals MSS.
Default Receive Window (RWIN) = 8760
RWIN Scaling (RFC1323) = 0 bits
Unscaled Receive Window = 8760

RWIN is a multiple of MSS
Other values for RWIN that might work well with your current MTU/MSS:
513920 (MSS x 44 * scale factor of 8)
256960 (MSS x 44 * scale factor of 4)
128480 (MSS x 44 * scale factor of 2)
64240 (MSS x 44)
bandwidth * delay product:

Your RcvWindow limits you to: 350.4 kbps (43.8 KBytes/s) @ 200ms
Your RcvWindow limits you to: 140.16 kbps (17.52 KBytes/s) @ 500ms
Consider increasing your RWIN value to optimize TCP/IP for broadband.
MTU Discovery (RFC1191) = ON
Time to live left = 121 hops

TTL value is ok.
Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON
IP type of service field (RFC1349) = 00000000

ps2wiz
07-10-03, 10:57 PM
By the way, Im using DSL (briged) caps are 1536/768. Im not using a router. And im using XP. My ISP is digizip which is on a Worldcom pipe (MCI/UNINET/Rythms).



HELP!

j79zlr
07-11-03, 01:00 AM
Download Cablenut (http://www.cablenut.com) and try these settings:

DefaultReceiveWindow = 196608
DefaultSendWindow = 98304
DisableAddressSharing = 1
InitialLargeBufferCount = 200
InitialMediumBufferCount = 480
InitialSmallBufferCount = 640
LargeBufferSize = 819200
MaxFastTransmit = 64000
MediumBufferSize = 150400
Priority Boost = 0
SmallBufferSize = 12800
TransmitWorker = 32
FastSendDatagramThreshhold = 4096
EnableFastRouteLookup = 1
EnablePMTUDiscovery = 1
IgnorePushBitsOnReceives = 0
GlobalMaxTcpWindowSize = 40880
MaxFreeTcbs = 8000
MaxHashTableSize = 16384
MaxNormLookupMemory = 5000000
SackOpts = 1
SynAttackProtect = 1
Tcp1323Opts = 0
TcpLogLevel = 1
TcpMaxDupAttacks = 3
TcpMaxHalfOpen = 100
TcpMaxHalfOpenRetired = 80
TcpRecvSegmentSize = 1460
TcpSendSegmentSize = 1460
TcpTimedWaitDelay = 30
TcpUseRFC1122UrgentPointer = 0
TcpWindowSize = 40880
MaxConnectionsPer1_0Server = 10
MaxConnectionsPerServer = 8
DefaultTimeToLive = 64
DefaultUserTOSSetting = 0
TcpMaxDataRetransmissions = 6
DefaultTOSValue = 90

ps2wiz
07-11-03, 01:27 AM
Internet Explorer game me these results


TCP options string = 020405b401010402
MTU = 1500
MTU is fully optimized for broadband.
MSS = 1460
Maximum useful data in each packet = 1460, which equals MSS.

Default Receive Window (RWIN) = 40880
RWIN Scaling (RFC1323) = 0 bits
Unscaled Receive Window = 40880

RWIN is a multiple of MSS
Other values for RWIN that might work well with your current MTU/MSS:
513920 (MSS x 44 * scale factor of 8)
256960 (MSS x 44 * scale factor of 4)
128480 (MSS x 44 * scale factor of 2)
64240 (MSS x 44)
bandwidth * delay product:

Your RcvWindow limits you to: 1635.2 kbps (204.4 KBytes/s) @ 200ms
Your RcvWindow limits you to: 654.08 kbps (81.76 KBytes/s) @ 500ms
MTU Discovery (RFC1191) = ON
Time to live left = 51 hops

TTL value is ok.
Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON
IP type of service field (RFC1349) = 01011010

Precedence (priority) = 010 (immediate)
Delay = 1 (low delay)
Throughput = 1 (high throughput)
Reliability = 0 (normal reliability)
Cost = 1 (low cost)





While mozilla still gave me the same old one. Does mozilla have to be tweaked seperatly or does that test not support mozila?

ps2wiz
07-11-03, 01:34 AM
Update, I have just ran a tweak test od dslreport, and this is what the results where:

http://monitor.dslreports.com/tweak/block:2c8e8fa?service=dsl&speed=1536&os=winXP&via=normal


Does speedguide's tcp-ip anlyzer not support mozilla? Also, it doenst like RWIN you gave. What do you think?

j79zlr
07-11-03, 02:22 AM
that looks good to me, how is your speed? The browser shouldn't make a difference, only thing I could think of is that it is in your cache still. Don't worry about DSLReports saying that the RWIN is out of range, the important thing from there is you had zero retransmissions.

ps2wiz
07-11-03, 03:09 AM
Im getting 1287/656 tops. Any other set-up that i could try/experiment with?

j79zlr
07-11-03, 03:12 AM
that looks pretty good, remember you aren't gonna get 100%. Anything 1200+ on a 1500 line is good and there is not much that can be done with upload, 656 is a good upload speed, wish I had that.

ps2wiz
07-11-03, 03:23 PM
Yes, I know thank you for the help so far. But like you guys say the goal is 90% and im only getting 84% of the upstream, im more concerned about it then the upstream.

ps2wiz
07-12-03, 07:33 PM
Can anyone recomend another tweak settings, maybe a new RWIN?

j79zlr
07-12-03, 08:12 PM
sorry, but there is very little that you can do about upload speed, RWIN affect the download, not upload

ps2wiz
07-12-03, 08:39 PM
Sorry, made a typo. Im more concerned bout the download speed. Any other recomendations now?

j79zlr
07-12-03, 09:23 PM
well you can find your personal best RWIN in a few steps. What you need to do is the following:

1) Start downloading a large file, like this one: ftp://ftp.newaol.com/aol6.0/179902/setupaol60.exe

2) Run the command, tracert www.yahoo.com

3) Take the average of the highest numbers that are similar, meaning if there are spikes of 350ms do not include them, but if most are around 150ms then that is your latency. This number should be between 100ms and 300ms.

4) Use that latency you found and enter your caps in here (http://home.covad.net/~zeiler07/cablenutXP2k.html)

mnosteele52
07-13-03, 08:39 AM
IMPORTANT TO UNDERSTAND
Remember your goal is to get about 90% of your rated speed, sometimes depending on factors beyond your control i.e. routers, servers, line quality, overloaded node etc. it just isn't gonna happen, no matter what you do.

ps2wiz
07-13-03, 10:50 PM
84%<90%

mnosteele52
07-13-03, 10:57 PM
Originally posted by mnosteele52
IMPORTANT TO UNDERSTANDsometimes depending on factors beyond your control i.e. routers, servers, line quality, overloaded node etc. it just isn't gonna happen, no matter what you do.

You obviously can't read.

:rolleyes: :rolleyes:

ps2wiz
07-14-03, 01:30 AM
Maybe if the font will get just a little bit bigger i might just make it out. :)