From: Pasi <pa...@ik...> - 2005-05-30 11:55:39
|
On Mon, May 30, 2005 at 02:28:31PM +0300, Pasi K=E4rkk=E4inen wrote: > On Wed, May 25, 2005 at 04:23:53PM +0200, Mark Martinec wrote: > > Pasi, > >=20 > > > I'm using amavisd-new with postfix. I need to bypass amavis based o= n sender > > > address/domain.. What's the best way to do this? > >=20 > > Possible with the help of Postfix configuration. > > > =20 > > > This needs to be done in postfix? how? > >=20 > > It is also possible to let Postfix assign a FILTER based on sender ad= dress, > > then use a dedicated policy bank in amavisd-new for such senders. > > See http://www.ijs.si/software/amavisd/amavisd-new-docs.html#pbanks > >=20 >=20 > I've been trying to figure out how to do this.. but haven't found solut= ion > yet. >=20 > I need to do this: >=20 > Disable virus scanning (or even whole amavis) for fo...@se... when h= e's > sending mail to ba...@re.... Encrypted zip's are blocked globally= , but > I need to allow some people send encrypted zips to some specific recipi= ents. >=20 > It seems that I cannot assign FILTER in header_checks in postfix to dis= able the=20 > content_filter, because the regexp matches are done per line..=20 > "From" and "To" fields are on different lines in the mail, so I cannot = match > both in the same "rule" using regexps. >=20 > I could only match the sender OR the recipient, and disable content_fil= ter > for them.. but that's too "wide open" system. >=20 > Any ideas? Thanks! >=20 Hmm.. would this be possible: In postfix, send all mails to port 10024 for amavis (default/global conte= nt-filter setting), but if sender matches regexp in postfix header_checks, assign d= ifferent=20 content-filter (port 10030 for example) for that mail. Then use amavis interface_policy/policy_bank to bypass scanning using bypass_virus_checks when mail comes in to amavis from port 10030.. ? Any comments on that? If I understood the documentation correctly, that would allow me to bypas= s scanning when "sender X is sending mail to recipient Y". - Pasi K=E4rkk=E4inen >=20 > > Mark > >=20 > >=20 |