dropping mechanism efficiency in bind server

Guy Pazi guy at wanwall.com
Thu May 31 11:19:28 UTC 2001


Hi,

I'm trying to get statistics on the efficiency of bind behavior under attack
of erroneous queries.

What is the ratio between handling a legitimate tcp/udp query and an
erroneous one? Are there specific sanity checks that cause an erroneous
query to consume more cycles then answering a legitimate one (such as bad
cksum, bad syntax, mismatch between declared and actual length, asking for
an unknown name)

Thanks
Guy



More information about the bind-users mailing list