*spamfilter* Channel options

From Messaging Server Technical Reference Wiki
Revision as of 10:09, 11 September 2017 by BulkPageCreator (Talk | contribs)

Jump to: navigation, search

destinationspamfilterN, destinationspamfilterNoptin, sourcespamfilterN, sourcespamfilterNoptin, disabledestinationspamfilterN, disablesourcespamfilterN Channel Options

The destinationspamfilterN, destinatationspamfilterNoptin, sourcespamfilterN, sourcespamfilterNoptin, disabledestinationspamfilterN, and disablesourcespamfilterN  channel options control whether spam/virus filter package processing is invoked by the MTA during message enqueue processing.

As of Messaging Server 7.0.5, the imexpire utility is also capable of invoking spam/virus filter packages, "as if" it were an MTA channel. For such use, imexpire is told a source channel as which to operate, and then any sourcespamfilterN and sourcespamfilterNoptin channel options for that channel specify what spam/virus filter package(s) to invoke.

The MTA supports the use of up to eight distinct spam/virus filtering packages, as configured via the spamfilterN_* MTA options (with N ranging between 1 and 8); the set of options with the same number all configure one of the distinct packages. And the number in a *spamfilterN* channel option correlates with which spam/virus filter package is to be invoked (or not invoked). (Note that the *brightmail* options, and the *spamfilter* options without an explicit number, are deprecated synonyms for the *spamfilter1* options. These deprecated forms are not allowed in a Unified Configuration.)

The *spamfilterNoptin channel options not only trigger spam/virus filter package processing, but do so with a particular "opt-in" value or values set, as some spam/virus filter packages (such as Brightmail) support different "choices" or "opt-in" values for what type of filtering will be performed (e.g., spam vs. virus). The MTA supports use of multiple, comma-separated "opt-in" values as an argument to an *optin channel option. Spam/virus filter package processing may also be triggered, with a specific "opt-in" value or values, via use of a per-user attribute (see the ldap_optinN and ldap_source_optinN MTA options) or a per-domain attribute (see the ldap_domain_attr_optinN MTA options). When "opt-in" values from multiple sources apply for a message -- for instance, if an "opt-in" value is set via a channel option as well as via an attribute---the MTA will pass all the applicable "opt-in" values to the spam/virus filter package.

The sourcespamfilter* channel options enable spam/virus filter package processing based upon source channel; when such a channel option applies, the spam/virus filter package will be activated towards the start of the SMTP dialogue (MAIL FROM stage). (For even earlier spam/virus filter package activation, see Spamfilter early verdicts.) The destinationspamfilter* channel options enable spam/virus filter package processing based upon destination channel, which is determined for each recipient at the RCPT TO stage of the SMTP dialogue; so if a destinationspamfilter* channel option applies, that spam/virus filter package activation occurs at RCPT TO stage.

The disable*spamfilterN channel options can be used to disable, on a channel-specific basic, spam/virus filtering that would otherwise be performed. In particular, note that the disable*spamfilterN channel options override any user-level opt-in to spam/filter package processing (as for instance selected via an LDAP attribute named by an ldap_optinN MTA option), and override any channel level opt-in (as for instance via a *spamfilterNoptin channel option). For instance, one might use disablesourcespamfilter1 on a tcp_auth channel to disable spam/virus filtering (by spam/virus filter package number 1) for all messages coming in the tcp_auth channel, overriding any spam/virus filter package use that might normally be triggered due to the destination channel or recipient address(es).

As of MS, the $+^ flag in the FROM_ACCESS mapping table can also be used to disable source channel spam filter optins. Also as of, the same $+^ flag in a recipient access mapping table can be used to disable any active spam filter optins at the current aliasing level. In either case the flag takes a comma separated list of spam filters to disable.

See also: