39559783   
   From: rick.jones2@hp.com   
      
   chris  wrote:   
   > Perhaps I wasn't clear.  The buffers on both ends are significantly   
   > larger than 24KB, and the quantity of in-flight data goes well above   
   > 24KB also.  24KB is just the size of the buffer-busting burst   
   > created by the sending TCP after a period of inactivity.   
      
   So, lots of buffering in the ends, only 24KB in the middle right?   
      
   > The application is capable of ramping up to a few Mb/s (with the   
   > previously described loss profile) before it chooses to stop, wait,   
   > spew, overflow, SACK, recover....  and back to steady state.   
      
   The receiver doesn't happen to advertise a zero window in there   
   anywhere does it?   
      
   rick jones   
      
   Not sure how much your setup would like TSO/LSO - Transport   
   Segmentation Offload/Large Send Offload - those can produce "bursts"   
   of perhaps as much as 64KB at a time... not sure that Windows 2003 has   
   support for that or not though - even if the NIC(s) involved do   
      
   --   
   I don't interest myself in "why". I think more often in terms of   
   "when", sometimes "where"; always "how much."  - Joubert   
   these opinions are mine, all mine; HP might not want them anyway... :)   
   feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)