GC is very expensive: am I doing something wrong?

Paul Rubin no.email at nospam.invalid
Wed Mar 24 02:39:49 CET 2010


Antoine Pitrou <solipsis at pitrou.net> writes:
>> See: http://www.cs.rice.edu/~scrosby/hash/ ...
> Certainly interesting in a purely academic point of view, but in real 
> life if you want to cause a denial of service by overwhelming a server, 
> there are far more obvious options than trying to guess the server's use 
> of hash tables and trying to cause lots of collisions in them.

If you look at the very low bandwidth used in some of those hashtable
attacks, it's hard to see any other somewhat-generic attack that's
comparably effective.  Usually we think of "DOS" as involving massive
botnets and the like, not a dribble of a few hundred characters per
second.



More information about the Python-list mailing list