Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 16 Feb 2005 18:30:46 -0800
From:      Julian Elischer <julian@elischer.org>
To:        net@freebsd.org
Subject:   Re: odd tcp trace.
Message-ID:  <421401D6.3040700@elischer.org>
In-Reply-To: <20050217021554.GD15240@yahoo-inc.com>
References:  <4213FD97.7000708@elischer.org> <20050217021554.GD15240@yahoo-inc.com>

next in thread | previous in thread | raw e-mail | index | archive | help


jayanth wrote:

>tcp large windows must have been negotiated which allows for window sizes > 32k.
>Look at the SYN packet and see if that is the case.
>
yes,
S 1555577485:1555577485(0) win 65535 <mss 1460,nop,wscale 1,nop,nop> (DF)
S 2834868684:2834868684(0) ack 1555577486 win 65535 <mss 1460,nop,wscale
2,nop,nop,> (DF)

19:53:49.692581 machineA.2293 > machineB.49177: . ack 1 win 33304
<nop,nop> (DF)
19:53:49.692610 machineB.49177 > machineA.2293: . ack 1 win 32942
<nop,nop> (DF)

but:
shouldn't tcpdump already be taking the scale into account?
I guess not, in which case the window is 64K but, then why did it stop
sending when the outstanding
data reached 33057?

>
>jayanth
>
>Julian Elischer (julian@elischer.org) wrote:
>  
>
>>This is a conversation between 2 FreeBSD 4.8 machines.
>>
>>Does anyone else worry about what is going on here regarding packets
>>sent and the received window?
>>
>>only odd setting here is that the slowstart flightsize was set to 30
>>(it was the only machine using this high latency link) (400mSec rtt but 
>>lots of bandwidth)
>>
>>19:53:49.694030 machineA.49177 > machineB.2293: . 28713:30161(1448) ack 
>>1 win 32942
>>19:53:49.694039 machineA.49177 > machineB.2293: . 30161:31609(1448) ack 
>>1 win 32942
>>19:53:49.694050 machineA.49177 > machineB.2293: . 31609:33057(1448) ack 
>>1 win 32942
>>

sender delays here when outstanding data reached ~ 32K. it has NOT
reached 30 packets yet..

>>19:53:50.069090 machineB.2293 > machineA.49177: . ack 2897 win 32580
>>19:53:50.069111 machineA.49177 > machineB.2293: . 33057:34505(1448) ack 
>>1 win 32942
>>19:53:50.069120 machineA.49177 > machineB.2293: . 34505:35953(1448) ack 
>>1 win 32942
>>19:53:50.069131 machineA.49177 > machineB.2293: . 35953:37401(1448) ack 
>>1 win 32942
>>19:53:50.069140 machineA.49177 > machineB.2293: . 37401:38849(1448) ack 
>>1 win 32942
>>19:53:50.069189 machineA.49177 > machineB.2293: . 38849:40297(1448) ack 
>>1 win 32942
>>19:53:50.069199 machineA.49177 > machineB.2293: . 40297:41745(1448) ack 
>>1 win 32942
>>19:53:50.069209 machineA.49177 > machineB.2293: . 41745:43193(1448) ack 
>>1 win 32942
>>19:53:50.069218 machineA.49177 > machineB.2293: . 43193:44641(1448) ack 
>>1 win 32942
>>19:53:50.069227 machineA.49177 > machineB.2293: . 44641:46089(1448) ack 
>>1 win 32942
>>19:53:50.069236 machineA.49177 > machineB.2293: . 46089:47537(1448) ack 
>>1 win 32942
>>19:53:50.069245 machineA.49177 > machineB.2293: . 47537:48985(1448) ack 
>>1 win 32942
>>
>>accrding to my calculations.. machine A should not fell it ok to send 
>>any packets beyond
>>sequence # 2897 + 32580
>>
>>yet it goes ahead and sends data up to 48985 which is WAY out of range..
>>
>>
>>_______________________________________________
>>freebsd-net@freebsd.org mailing list
>>http://lists.freebsd.org/mailman/listinfo/freebsd-net
>>To unsubscribe, send any mail to "freebsd-net-unsubscribe@freebsd.org"
>>
>>
>>    
>>




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?421401D6.3040700>