Lab test: Barracuda Spam Firewall

Barracuda combines solid features and great configurability at a low price, but filtering accuracy lags the field

The Barracuda Spam Firewall blocked more spam before filtering than any other appliance, using IP address reputation. As a result, spam as a percentage of total messages received was the lowest in the test. However, Barracuda also had the lowest percentage of spam caught, 88 percent, which is acceptable but not great. Lower accuracy sometimes helps reduce the number of false positives, but it didn’t seem to help the Barracuda much: one critical false positive and 33 bulk false positives rank sixth and fifth, respectively, among the nine solutions tested. Nevertheless, in terms of overall filtering performance, the Barracuda is definitely usable, especially after bulk senders are whitelisted.

The Barracuda model 400 and all of its brethren, which differ only in number of users supported, is easy to set up, although there are a great many settings that can be changed from the defaults, giving you great flexibility in configuration. You can create a default IP address without serial or KVM login, which makes the initial network setup easier, and the rest of the configuration is done through a Web browser. LDAP setup isn’t the easiest, but it's not too bad.

There are many default settings that can be tailored to match your requirements, although how the settings will affect performance isn’t always clear. The default settings and the settings Barracuda recommended to me were different, so the product may require tuning to optimize performance.

[ Compare spam filtering statistics for the mail security appliances tested. Compare their features. Return to “Test Center guide: Mail security appliances.” See reviews of the appliances from Barracuda, BorderWare, Cisco, Mirapoint, Proofpoint, Secure Computing, Sendio, Symantec, and Tumbleweed. ]

The default settings also block a substantial amount of legitimate bulk e-mail. According to Barracuda, many of these legitimate messages may not comply with the CAN-SPAM act. To clear the way for these messages to get through, you’ll need to allow for a training period during which users whitelist the stuff they want.

Using the Web interface to access the quarantine is pretty clumsy. There are three separate steps to releasing blocked messages: whitelisting the sender, marking the message as ‘not spam’ for re-classification, and delivering the message. Each step requires a box to be checked, and after the screen refreshes, the first box doesn’t stay marked, so you have to wait a few seconds for the refresh, find the box again, check again, click a second button, wait again, find it again then click a third button. Thankfully, using the Outlook plugin makes whitelisting much easier.

Adding senders to the whitelist adds the sender only, with no option to add the entire domain. This can make for a lot of extra work, because senders are often different for every newsletter even though they’re all from the same domain.

Barracuda offers a good spread of additional features, including compliance filtering, IM protection, URL filtering to help fight phishing, clustering for higher loads, scanning of attachments as well as the body of e-mails, and good reporting.

The Barracuda has the lowest price by a substantial margin of any product tested, and because the software is the same for all versions of the appliance, upgrading to a new version that supports more users would be painless as well as inexpensive. Ranging from the model 100 that supports 1 to 50 users at a price of $899, up to the model 900 that supports 15,000 to 30,000 users at a price of $29,999, the Barracuda Spam Firewall offers a very low cost per user and good functionality.

InfoWorld Scorecard
Effectiveness (30.0%)
Setup (10.0%)
Manageability (30.0%)
Scalability (20.0%)
Value (10.0%)
Overall Score (100%)
Barracuda Spam Firewall 400 v3.4.10.102 7.0 8.0 7.0 8.0 9.0 7.5
Mobile Security Insider: iOS vs. Android vs. BlackBerry vs. Windows Phone
Recommended
Join the discussion
Be the first to comment on this article. Our Commenting Policies