Re: [tcpm] RFC 2581 Clarification

[Yoshifumi Nishida <[e-mail]>]

Hello Leino,

I think you had better refer rfc3517 instead of rfc2581 for this.

Thanks,
--
Yoshifumi Nishida
[e-mail]

From: "Leino, Tammy" <[e-mail]>
Subject: [tcpm] RFC 2581 Clarification
Date: Wed, 1 Apr 2009 14:29:06 -0600
Message-ID: <[e-mail]>

 > Hello All,
 > 
 >  
 > 
 > Can someone please clarify what is meant by "a segment" in step 4 of
 > section 3.2 when discussing Fast Recovery:
 > 
 >  
 > 
 >    4.  Transmit a segment, if allowed by the new value of cwnd and the
 >        receiver's advertised window.
 > 
 >  
 > 
 > Specifically, could this "segment" be data on the retransmission queue
 > that has not been ACKed (or retransmitted by Fast Retransmit) or is this
 > "segment" only new data that has not yet been queued for retransmission?
 > 
 >  
 > 
 > I had always interpreted this as new data that has not been queued for
 > retransmission, but in reading RFC 2018 regarding SACK, it seems that
 > data on the retransmission queue is expected to be retransmitted in Fast
 > Recovery as this section is outside the scope of the retransmission
 > timer:
 > 
 >  
 >    After the SACKed bit is turned on (as the result of processing a
 >    received SACK option), the data sender will skip that segment during
 >    any later retransmission. 
 > 
 >  
 > 
 > Your insight is greatly appreciated.
 > 
 >  
 > 
 > Best Regards,
 > Tammy Leino
 > 
 >  
 >