I don't get it

Mohan-san · Post by **Mohan-san** » Wed Aug 21, 2002 9:17 am

I changed my Rwin to a random number, and not being a multiple of MSS (1460) my speeds have gone up.

Test running..........
** Speed 470(down)/231(up) kbps **
(At least 9 times faster than a 56k modem)
Finish.

Test running..........
** Speed 474(down)/230(up) kbps **
(At least 9 times faster than a 56k modem)
Finish.

Test running..........
** Speed 476(down)/231(up) kbps **
(At least 9 times faster than a 56k modem)
Finish.

Test running..........
** Speed 477(down)/247(up) kbps **
(At least 9 times faster than a 56k modem)
Finish.

They used to be much lower, infact a bit like this

Test running..........
** Speed 420(down)/230(up) kbps **
(At least _ times faster than a 56k modem)
Finish.

mnosteele52 · Post by **mnosteele52** » Wed Aug 21, 2002 9:35 am

Just because it's not a mulptiple doesn't mean it won't work, but if it is a multiple is reduces the chance for lost packets thereby slowing you down. Try taking the number you used and rasie it ir lower to make it a multiple.

From the White Papers:

.............TCP adjusts to even increments of the maximum segment size (MSS) negotiated during connection setup. Matching the receive window to even increments of the MSS increases the percentage of full-sized TCP segments used during bulk data transmission.

Mohan-san · Post by **Mohan-san** » Wed Aug 21, 2002 10:04 am

Welll its 27670 right now. So I'll change it to 27740. Let's see...

Mohan-san · Post by **Mohan-san** » Wed Aug 21, 2002 10:17 am

Right..

Test running..........
** Speed 474(down)/231(up) kbps **
(At least 9 times faster than a 56k modem)
Finish.

Test running..........
** Speed 472(down)/210(up) kbps **
(At least 9 times faster than a 56k modem)
Finish.

Test running..........
** Speed 477(down)/213(up) kbps **
(At least 9 times faster than a 56k modem)
Finish.

Test running..........
** Speed 477(down)/242(up) kbps **
(At least 9 times faster than a 56k modem)
Finish.

See any difference?

mnosteele52 · Post by **mnosteele52** » Wed Aug 21, 2002 10:29 am

I think your missing the point. It increases the percentage of receiving more full sized data packets.

TcpWindowSize

Description: This parameter determines the maximum TCP receive window size offered by the system. The receive window specifies the number of bytes a sender may transmit without receiving an acknowledgment. In general, larger receive windows will improve performance over high (delay * bandwidth ) networks. For highest efficiency, the receive window should be an even multiple of the TCP Maximum Segment Size (MSS).

Mohan-san · Post by **Mohan-san** » Wed Aug 21, 2002 11:54 am

Ahh I see.

vicoman · Post by **vicoman** » Wed Aug 21, 2002 5:57 pm

Some tweaking programs and even speed patches from other sites, do not always use a multiple of MSS. What I would like to know is what is the formula they use to come up with these numbers or are they just pulling them out of the air, which I doubt. Does anyone have a link to a informative page on this subject? I haven't been able to find anything.

jeff8874 · Post by **jeff8874** » Wed Aug 21, 2002 6:03 pm

I have tested almost evey possible RWIN and I find that the best that works for me is 372296, which is not a multiple obviously. Xsetup changed it to that and I never felt the need to change it back. It works great.