Rate limits: Difference between revisions

From ASRG
Jump to navigationJump to search
(New page: {{ast |date=early 2000 |difficult=Low |popular=High |effective=Low |harm=Low |where=MTA }} The idea behind rate limit is that usually humans and legitimate mail servers send messages at s...)
 
No edit summary
 
Line 7: Line 7:
|where=MTA
|where=MTA
}}
}}
Humans and legitimate mail servers usually send messages at a limited  rate. Spam robots can, sometimes, send bursts of messages. An SMTP server can count the number of connections per client over some time window and reject connections, with a temporary reply code, when a threshold is reached. Limits are usually applied to connections, messages or recipients.


The idea behind rate limit is that usually humans and legitimate mail servers send messages at some limited or small rate. Robots can, sometimes, send bursts of messages. So, the SMTP server shall evaluate the number of connections, per client, over some time window and reject connections, with a temporary reply code, when some threshold is reached. Limits are usually applied to connections, messages or recipients.
Nowadays, rate limiting effectiveness is too low to be considered as a real anti-spam measure, but it remains very effective against DoS and other abuses. It can suffer from false positives against legitimate bursty mail sources such as mailing lists.
 
Nowadays, rate limiting effectiveness is too low to be considered as a real anti-spam measure, but it remains very effective against DoS and other abuses.

Latest revision as of 18:11, 30 May 2009

Anti-spam technique: Rate limits
Date of first use: early 2000
Effectiveness: Low
Popularity: High
Difficulty of implementation: Low
Where implemented: MTA
Harm: Low

Humans and legitimate mail servers usually send messages at a limited rate. Spam robots can, sometimes, send bursts of messages. An SMTP server can count the number of connections per client over some time window and reject connections, with a temporary reply code, when a threshold is reached. Limits are usually applied to connections, messages or recipients.

Nowadays, rate limiting effectiveness is too low to be considered as a real anti-spam measure, but it remains very effective against DoS and other abuses. It can suffer from false positives against legitimate bursty mail sources such as mailing lists.