And since HTTP is an interactive protocol, your users would be up in arms about how long it takes to scan when they go to browse a page; Emails, who cares if there's a delay of 30 seconds? (unless you're using Email for real-time messaging, which you shouldn't...)
Some people kvetch about the one or two second delay that an HTTP proxy takes! Now picture a five to ten second delay (and that would be an optimistic estimate...)
Hi secapp, a good coded virusscanner combined with a right configured http-proxy on a right sized computer should not have such a delay. Do you mean that a virusscanner takes 3 to 8 seconds to scan a single side?
We've used such products on CheckPoint and the performance was atrocious; to identify virile patterns it had to first cache the entire file, then consult a pattern database; the users were screaming.
For such protection I sooner rely on centrally managed browser security settings with a centralized list of trusted websites; all those sites not trusted run in high security mode; a request can be made to knowledgeable people when users want a site added to the list...
Not to mention the fact that virus scanners won't catch numerous malicious one-off variants...
We've used such products on CheckPoint and the performance was atrocious; to identify virile patterns it had to first cache the entire file, then consult a pattern database; the users were screaming.
For such protection I sooner rely on centrally managed browser security settings with a centralized list of trusted websites; all those sites not trusted run in high security mode; a request can be made to knowledgeable people when users want a site added to the list...
Not to mention the fact that virus scanners won't catch numerous malicious one-off variants...
