From: grahn+nntp@snipabacken.se   
      
   On Mon, 2010-11-08, Rick Jones wrote:   
   > Jorgen Grahn  wrote:   
   >> - Are you allowed to put extra load on the server, stealing its CPU   
   >>   resources or SIGSTOP the server software momentarily?   
   >   
   > Might be OK for simulating a very heavily CPU loaded server, but I'd   
   > avoid it for simulating a heavily loaded network.  I was involved in a   
   > benchmark that was doing something conceptually similar - not calling   
   > recv() except at a rate based on the presumed simulated bandwidth -   
   > and the zero window situations were inducing TCP behaviours a realy   
   > link of that bandwidth would not have induced.  I'm reasonably   
   > confident that SIGSTOP would do the same thing.   
      
   You're probably right -- I was just enumerating a number of things   
   which came to mind. Anything is better than testing on two idling   
   overpowered servers connected with ten feet of 1Gbit Ethernet ...   
   I've done that too often myself.   
      
   /Jorgen   
      
   --   
     // Jorgen Grahn    O  o   .   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)