Home | Sitemap | Recent Changes | Login

SPF Logo

Sender Policy Framework

History of FAQ/DDoS


Revision 3 . . (edit) 2006-12-10 16:06 (UTC) by Julian Mehnle
Revision 2 . . 2006-12-10 16:05 (UTC) by Julian Mehnle
Revision 1 . . 2006-03-11 18:47 (UTC) by Scott Kitterman
  

Difference (from prior major revision) (minor diff)
Paragraph 1Paragraph 1

== Given a large enough spam run, SPF becomes a DDOS against the forged domain!

== Given a large enough spam run, SPF becomes a DDoS attack against the forged domain!



Each SMTPd in a spam run may send a DNS query to the forged host's nameservers. At a million SMTPds, that's 100 megabytes of traffic!

Each SMTP MTA in a spam run may send a DNS query to the forged host's nameservers.  At a million MTAs, that's 100 megabytes of traffic!



DNS queries are still smaller than bounce messages. And most SPF lookups can be cached; only the relatively uncommon [http://www.openspf.org/mechanisms.html#exists exists] mechanism doesn't benefit from caching.

DNS queries are still smaller than bounce messages.  And most SPF lookups can be cached; only the relatively uncommon "<tt>[[SPF Record Syntax#exists|exists]]</tt>" mechanism doesn't benefit from caching (because it is usually used with macros, which make caching difficult).



Strict [http://new.openspf.org/blobs/draft-schlitt-spf-classic-02.html#anchor31 processing limits] have been put in place to mitigate the risks associated with DNS loading from SPF.

Strict [[RFC 4408#processing-limits|processing limits]] have been put in place to mitigate the risks associated with DNS loading from SPF.