MPP v3

February 4, 2008

Integrated Postfix Policy Server

We are excited to announce that MPP now features an integrated Postfix Policy Server. Our Postfix policy protocol allows Postfix to delegate access decisions to a policy server during the SMTP transaction.
Historically, MPP has been a post-queue content filter, but with this upgrade MPP can now
control whether an email is rejected pre-queue or processed post-queue. This capability, which intelligently decides between pre or post queue filterings based on the characteristics of the email, easily makes MPP the most versatile content filter for Postfix available.

Also, MPP’s powerful policy engine enables the configuration of a great number of our features on a per-policy basis. If an email arrives for a single recipient and the email has been blacklisted by the group policy or receiver, the email can be rejected pre-queue, but if a multi-recipient email arrives with two policy matches, then the email must be accepted and processed post-queue.

No other Postfix filter that we know of has this superior capability.

GUI Redesign
The MPP GUI has been completely redesigned with a new look and reorganized controls giving it a more intuitive and intelligent functionality.

Body Stripping
MPP can strip any MIME part of an email, place the content on an ftp or http server, and replace the original MIME part with a link to the file. MPP has multiple logic criteria to match based on filename, size, type, etc.

This is useful for reducing mailbox sizes by centrally storing attachments and also gives a company much better control over their confidential documents.

Thresholds
MPP can detect and report users that send too much email or too much spam in specified intervals.

Sun Java Systems Messaging Server
MPP now supports the Sun Java Systems Messaging Server, an email server that is used by many large service providers.

Logging Enhancements
MPP supports syslog for centralized logging. For our proprietary, integrated logs MPP supports splitting log files based on size, time or on reload.

Per-Policy Group RBL’s
MPP now supports real-time black hole lists on a per-policy basis. When combined with our Postfix Policy Server MPP can easily enforce per-domain SMTP policy.

New Platforms, Operating Systems and Distributions
MPP now supports FreeBSD6, Mac Intel platforms, Debian, and we have introduced new builds for newer distributions of Linux that use GCC greater than 3.3. Also, installation scripts have been significantly overhauled to ease installation.

LMTP Support
MPP now supports LMTP for per-user status for Postfix, Exim and Sun JSMS.

Minor Enhancements
Defer action for scanning error, qmail connection timers, IP identitity extracted from qmail and sendmail,

MPP v3.6

February 4, 2008

MPP 3.6 Feature List

  • Cloudmark 2 API implemented

  • MySQL code rewritten and new MySQL schema has been made default. New SQL scheme
    provides better scalability for multi-domain environments and facilitates distributed
    databases.

  • Qreview v3 is included in installation package. QReview provides web based access to
    spam quarantines and email archives.

  • Added defer action for Thresholds / Autoblacklists. This feature introduces rate limiting,
    as an abusive sender will no longer have their email blocked, but will see
    an SMTP defer if Postfix Policy Server is in use, or their email will be deferred
    in the mail queue.

  • Access Control Lists are made case-insensitive

  • Increased logging information by adding missing codes to decision path in log.

  • Configurable sender and recipient(s) for threshold alerts. Thresholds are useful
    to identify outbound spammers, heavy inbound senders or to automatically black list
    repeat, high volume spammers. Traditionally threshold alerts were only sent to the
    postmaster, now the admin can set the address for alerts.

  • Automatic Client Host Black Listing of spam trap offenders. Spam traps are
    regular expression based template addresses that are set for spammers. If
    spammers hit the templates their IP addresses are automatically added to a black list.
    Traditionally MPP spamtrap offenders were blocked on a per-group basis, now administrators
    can choose to block them globally for all policy groups. This increases efficiency
    in MPP processing and causes an automatic disconnect of spammers. A single spam trap
    can block hundreds of thousands of dictionary attacks email in a day.

  • MPP has an incredibly powerful Postfix Policy Server and 3.6 offers Postfix olicy Server improvements.
    3.6 offers better support of a standalone policy server by sending OK is sent instead of FILTER if
    standalone mode is in use. 3.6 increases scalability by settings policy_threads_max to
    400 by default and is limiting max threads to 400. Interoperability is improved
    by changing OK response to DUNNO for Policy Server.

Please see the release notes for more detailed information about bug fixes and minor updates

Previous Releases

MPP v3.4

February 4, 2008

MPPv34 addresses issues of scalability, functionality and usability.

Feature List

  • Dynamic threading for Postfix Policy Server - Automatically create and destroy threads based on load.
  • Internal disk queue for quarantine and archival - Remove direct writing to DB under congestion.
  • Hierarchical quarantine/archival - MPP can use esmtp to transport quarantine/archive traffic to a centralized instance of MPP that then writes to DB
  • International charachter support for signatures/disclaimers
  • Loading of mppd.conf.xml from stdout, we now provide a script to load mppd.conf.xml from a tftp server and send it to stdout for mppd to use.
  • Custom Reject Notices and API - Write custom reject notices or use MPP as standalone scanning service that can be implemented in scripts.
  • 2 new macros for maildir storage and small maildir fixes. Macros are for first/second letter for user part/domain part of mail address: %U1%, %U2%, %D1%, %D2%.
  • Whitelists by IP for specific features of greylists, spam traps and tresholds/auto-blacklists
  • Custom spam scoring - Assign scores to RBL sites and spam engines as an alternative to first-match actions
  • McAfee UVScan command line scanner is now supported
  • Improved MySQL formats for large scale spam quarantine