[RFC] simulating network properties for benchmarks

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Carl Friedrich Bolz wrote:
>

...

>>> The disadvantage is that lsprof won't understand what's going on-- but
>>> does sleep show up in lsprof, or does it measure CPU time?
>> I'm not sure.   I thought it was wall time.
> 
> I think so too, yes. I think it would indeed be useful to track the
> sleeps with lsprof.
> 
> Cheers,
> 
> Carl Friedrich

I'm pretty sure it is wall clock time, because you can do:

bzr --lsprof pull

And it will tell you how much time was spent waiting on the socket.
Which I think could be something that we want to know about in a real
benchmark test.

Especially if we get a better ability to do pipelined requests or async.
At that point we may be able to do a lot while we are waiting for a
request to return, and so we will need a real sleep, rather than a fake one.

But maybe it's best to have a parameterized sleep, because for now, a
fake sleep is as good as a real one, and it does make the test finish
faster.

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEwNoJJdeBCYSNAAMRAmmXAJ9buJfJ1+4rX/IvnnSbTpBHEqDFpwCeMV/d
LeBv6T4gZ7g/MznnxApDJd4=
=k1th
-----END PGP SIGNATURE-----