SPF, MTAs and SRS
Sender Policy Framework (SPF) takes aim at the practice of return-path spoofing, a technique employed by worms, viruses and other senders of unwanted mail. SPF consists of two parts. First, domain administrators publish SPF records in the DNS. Those records describe the servers the domain uses for outbound mail. Then they are read by SPF-enabled MTAs. Mail coming from a server not described in SPF can be considered forged.
This article, the second of a two-part series, explains how to add SPF capabilities to your mail server. It also discusses how e-mail forwarding and Web-generated e-mail services can adjust to SPF by performing sender rewriting.
This article was written in early February 2004 and reflects the state of the Internet at that time. The MyDoom worm, a virus that spoofs return-path addresses, recently had littered millions of mailboxes with bogus bounce messages.
Last month I described how to construct an SPF record, and DNS administrators all over the world responded. First, they published records, then crossed their fingers and waited. What are they waiting for? They're waiting for you. They've made it possible for you to distinguish their legitimate mail easily from forgeries. Now it's your turn to help them cut down on bogus bounces and abuse reports. If you're just tuning in, see the on-line Resources section for last month's article and an easy Web-based SPF wizard.
The major mail transfer agents (MTAs) in the Linux world are Sendmail, Postfix, Qmail and Exim. Although most antispam vendors already have SPF support included in their products or plan to add it in their next release, MTAs tend to want to leave that task up to you. Most MTAs offer an interface into which you can plug your antispam tool.
SPF can be made to work in your MTA in two ways. If you're the kind of sysadmin who prefers to compile your own software, start at the SPF downloads page. There you can find the SPF plugin that's right for your MTA, plus detailed installation instructions. If you prefer to manage your software using a package system, you may find an SPF-enabled version of your MTA already built and ready to install.
Most of the plugins rely on the reference Perl library Mail::SPF::Query. You can install that library directly from CPAN, or you can try to find a package for it. It provides a simple program to run SPF queries at the command line. It also provides a simple dæmon that handles SPF query requests over a UNIX domain or inet socket.
By default, most of the plugins tell the MTA to reject messages that fail SPF tests and add a Received-SPF header to the rest. Conservative installations may prefer to add the line Received-SPF: fail instead of rejecting. This configuration option is described in the plugin documentation.
Sendmail's plugin interface is called Milter (see on-line Resources). Recent Sendmail versions have Milter capability compiled in by default. Sendmail talks to Milter through a socket interface. Sendmail tells Milter about the incoming SMTP transaction, and Milter tells Sendmail what to do. Milter runs as a dæmon and needs to be started separately.
Two Milters should be available at the SPF Web site: one in Perl and one in C. The Perl version is a little more mature, but if you need speed, the C version may be a better choice.
To make Milter work with Sendmail, add a couple of lines to your sendmail.mc file, rebuild sendmail.cf and restart Sendmail.
If you'd rather not use Milter, libspf comes with a patch that integrates SPF directly into Sendmail.
Postfix 2.1 comes with a policy dæmon interface. It works much like Milter does: Postfix connects to the dæmon and provides a play-by-play commentary, and the dæmon returns an action to Postfix. If you're running a recent development snapshot of version 2.0, make sure you're using 2.0.18-20040122 or later.
Policy dæmons are configured in main.cf and master.cf. They are managed by Postfix, which starts and stops them as needed, so you don't need to worry about that. The Postfix policy dæmon is written in Perl and calls the standard Mail::SPF::Query library.
Exim 4 introduced Access Control Lists (ACLs), a powerful and compact mini-language for making antispam and other local policy decisions. The ACL code that handles SPF for Exim is only about 12 lines long.
You need to install the Mail::SPF::Query library and run its SPF dæmon, which listens on a socket. The SPF ACL connects to the spfd and reads it the client IP, HELO argument and MAIL FROM sender address. It then receives an SPF result, a response for the SMTP server and a Received-SPF header line. You need to start the spfd separately.
|Speed Up Your Web Site with Varnish||Jun 19, 2013|
|Non-Linux FOSS: libnotify, OS X Style||Jun 18, 2013|
|Containers—Not Virtual Machines—Are the Future Cloud||Jun 17, 2013|
|Lock-Free Multi-Producer Multi-Consumer Queue on Ring Buffer||Jun 12, 2013|
|Weechat, Irssi's Little Brother||Jun 11, 2013|
|One Tail Just Isn't Enough||Jun 07, 2013|
- Containers—Not Virtual Machines—Are the Future Cloud
- Non-Linux FOSS: libnotify, OS X Style
- Linux Systems Administrator
- Lock-Free Multi-Producer Multi-Consumer Queue on Ring Buffer
- Validate an E-Mail Address with PHP, the Right Way
- Technical Support Rep
- Senior Perl Developer
- UX Designer
- Introduction to MapReduce with Hadoop on Linux
Free Webinar: Hadoop
How to Build an Optimal Hadoop Cluster to Store and Maintain Unlimited Amounts of Data Using Microservers
Realizing the promise of Apache® Hadoop® requires the effective deployment of compute, memory, storage and networking to achieve optimal results. With its flexibility and multitude of options, it is easy to over or under provision the server infrastructure, resulting in poor performance and high TCO. Join us for an in depth, technical discussion with industry experts from leading Hadoop and server companies who will provide insights into the key considerations for designing and deploying an optimal Hadoop cluster.
Some of key questions to be discussed are:
- What is the “typical” Hadoop cluster and what should be installed on the different machine types?
- Why should you consider the typical workload patterns when making your hardware decisions?
- Are all microservers created equal for Hadoop deployments?
- How do I plan for expansion if I require more compute, memory, storage or networking?