This is a static archive of our old Q&A Site. Please post any new questions and answers at ask.wireshark.org.

TCP receive window

0

Hello,

I am trying to undestand TCP receive window. For example, on a iperf test, if I plot the TCP Time-Sequence Graph selecting any packet from client to server (remember iperf makes the test moving data from client to server), i see the above line looks like it is limiting the sender to send more data. But how that info can be found from a client to server packet? If I am not wrong, the "windows size" is the RWIN size, which means the amount of data that can be "received" by the client, which is not what we want to test. Am I correct?

thanks!

asked 26 Apr '15, 00:26

luicson's gravatar image

luicson
6112
accept rate: 0%

converted to question 26 Apr '15, 05:21

grahamb's gravatar image

grahamb ♦
19.8k330206


2 Answers:

1

RWIN is the amount of data that can be received by the receiver, meaning: the node that gets the data. It doesn't matter what you call a client or a server; both have window sizes. The window size of one is the amount the other can send without having to wait for an ACK, and vice versa. So if your "client" is sending data to the "server" it's the server's RWIN that matters.

answered 27 Apr '15, 07:25

Jasper's gravatar image

Jasper ♦♦
23.8k551284
accept rate: 18%

0

If your question is about TCP Window size and Flow control, I think this section from the TCP/IP Guide should provide you the best answer:

http://www.tcpipguide.com/free/t_TCPWindowSizeAdjustmentandFlowControl-2.htm

answered 27 Apr '15, 06:04

Amato_C's gravatar image

Amato_C
1.1k142032
accept rate: 14%