Page 1 of 1
changing RWIN scaling
Posted: Wed Jun 06, 2001 10:42 pm
by bigjohns97
where is the reg entry to change my rwin scaling to 4
TCP options string = 020405b40103030201010402
MTU = 1500
MTU is fully optimized for broadband.
MSS = 1460
Maximum useful data in each packet = 1460, which is equal to MSS.
Default Receive Window (RWIN) = 256960
RWIN Scaling (RFC1323) = 2 bits
Unscaled Receive Window = 64240
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)
128480 (MSS x 44 * scale factor of 2)
64240 (MSS x 44)
bandwidth * delay product:
Your RcvWindow limits you to: 10278.4 kbps (1284.8 KBytes/s) @ 200ms
Your RcvWindow limits you to: 4111.36 kbps (513.92 KBytes/s) @ 500ms
MTU Discovery (RFC1191) = ON
Time to live left = 46 hops
TTL value is ok.
Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON
IP type of service field (RFC1349)= 00000000
thanks in advance!
Posted: Wed Jun 06, 2001 10:44 pm
by EvilAjax
Scaling can only be enabled. Tcp1323Opts = 1. The scaling changes with a higher RWIN. Me thinks
Posted: Wed Jun 06, 2001 10:46 pm
by EvilAjax
What kind of a connection do you have? Can I have your system spex? And what are your advertised speeds?
Welcome to speedguide.
Posted: Wed Jun 06, 2001 10:48 pm
by bigjohns97
i have a p3 800 on an abit sh6 with 384 megs of ram in austin tx on road runner with no tweaks enabled except the default winme tweak on speedguide.
Posted: Wed Jun 06, 2001 10:52 pm
by bigjohns97
Originally posted by EvilAjax
Scaling can only be enabled. Tcp1323Opts = 1. The scaling changes with a higher RWIN. Me thinks
If i am not mistaken my scaling is already enabled i need it to read 4 when i do this analyzer?
Really?
Posted: Wed Jun 06, 2001 11:41 pm
by iaus10
If you enter a RWIN of 642400, you're scaling will be at 4. Probably not worth it... Lots of packet loss.
Posted: Wed Jun 06, 2001 11:46 pm
by cablenut
"If you enter a RWIN of 642400, you're scaling will be at 4. Probably not worth it... Lots of packet loss"
THis is a misconception amoung alot of people. First of all we can't test our RWIN's with ping because ping uses ICMP which does not use TCP options or RWIN's. Second 99% of the time any RWIN over 64240 is not used and is scaled down.
If he uses any of the scaling factors (2,4,6, or 8) the RWIN will be scaled down correctly.
Posted: Thu Jun 07, 2001 12:12 am
by iaus10
Cablenut,
Are we talking about the same thing? Thanks for the info. By the way, I don't normally use this RWIN, just trying to see what bigjohns97 was goin' for.
MTU = 1500
MTU is fully optimized for broadband.
MSS = 1460
Maximum useful data in each packet = 1460, which is equal to MSS.
Default Receive Window (RWIN) = 642400
RWIN Scaling (RFC1323) = 4 bits
Unscaled Receive Window = 40150
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: 25696 kbps (3212 KBytes/s) @ 200ms
Your RcvWindow limits you to: 10278.4 kbps (1284.8 KBytes/s) @ 500ms
MTU Discovery (RFC1191) = ON
Time to live left = 32 hops
TTL value is ok.
Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON
IP type of service field (RFC1349)= 00000000
Posted: Thu Jun 07, 2001 9:21 am
by bigjohns97
Originally posted by iaus10
Cablenut,
Are we talking about the same thing? Thanks for the info. By the way, I don't normally use this RWIN, just trying to see what bigjohns97 was goin' for.
MTU = 1500
MTU is fully optimized for broadband.
MSS = 1460
Maximum useful data in each packet = 1460, which is equal to MSS.
Default Receive Window (RWIN) = 642400
RWIN Scaling (RFC1323) = 4 bits
Unscaled Receive Window = 40150
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: 25696 kbps (3212 KBytes/s) @ 200ms
Your RcvWindow limits you to: 10278.4 kbps (1284.8 KBytes/s) @ 500ms
MTU Discovery (RFC1191) = ON
Time to live left = 32 hops
TTL value is ok.
Timestamps (RFC1323) = OFF
Selective Acknowledgements (RFC2018) = ON
IP type of service field (RFC1349)= 00000000
This is what confuses me
Your current rwin is 642400 which is way over the recommended rwin from speedguide. Yet your scaling is only at 4 when this table states "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)"
You see what i am getting at for your scaling to only be at 4 with rwin at 642k then your actual mssx44 is wrong, wether it's the actuall mss or the 44 that is changed i am not sure but it seems to me by this table that it's the only thing that is constant. Constant meaning that has been tested and is known for experiementation as a constant. what i would like to know is how they (table) get an rwin of 256960 scale 4 , or 513920 scale 8, those are the numbers i am looking for when i run this analyser.
Thanks for any help with this question of mine.
Posted: Thu Jun 07, 2001 9:29 am
by Lobo
bigjohns97, try RWIN's of 513920, 522680 or 524176
Posted: Thu Jun 07, 2001 9:35 am
by iaus10
Hey Lobo, what is the deal with "RWIN Scaling (RFC1323)" and how does it relate to MSS?
Posted: Thu Jun 07, 2001 9:42 am
by Lobo
Only MTU relates to MSS
Posted: Thu Jun 07, 2001 9:45 am
by iaus10
Yeah... but how does the scaling get calculated? 1 bit, 2bits, 3bits, etc.
Posted: Thu Jun 07, 2001 9:49 am
by Lobo
By how big your RWIN is, MTU and MSS stay the same
Posted: Thu Jun 07, 2001 10:01 am
by iaus10
Got it, but here's how I'm confused: What is the mathematical relationship between an RWIN = 256960, and the RWIN Scaling (RFC1323) = 2 bits? Or is there no relationship other than the RWIN Scaling (RFC1323) being a set of parameters that is used?
Posted: Thu Jun 07, 2001 10:05 am
by bigjohns97
Originally posted by Lobo
By how big your RWIN is, MTU and MSS stay the same
Or does 4bits=a scale factor of 8???????
Posted: Thu Jun 07, 2001 10:07 am
by Lobo
Ask Phillip or Cablenut these hard ones
Posted: Thu Jun 07, 2001 10:13 am
by iaus10
scale factor of 8 (513920) shows an RWIN scaling of 3bits. Confused
Phillip... Cablenut ?
Posted: Thu Jun 07, 2001 10:43 am
by Lobo
Or Dannjr could answer this
or Brent or rmrucker, Kip Patterson, Storm90, well big list
Posted: Thu Jun 07, 2001 12:19 pm
by bigjohns97
Originally posted by Lobo
Or Dannjr could answer this
or Brent or rmrucker, Kip Patterson, Storm90, well big list
Can someone take this to the cablenut fourms , i cant register on there while i'm at work for some reason.
Posted: Thu Jun 07, 2001 1:56 pm
by Kip Patterson
The scaling factor refers to how many digits (binary) that the RWIN is multiplied by. That's what the word "bit" meant originally - "a contraction of "binary digit". If the scaling factor is 4 bits, that is 2x2x2x2 or 16. If you look at the numbers posted earlier, you'll see that the RWIN is 16x the unscaled RWIN.
Kip