Energize Updates
To provide maximum protection against the latest types of spam and virus attacks, Barracuda Networks maintains a powerful operations center called Barracuda Central. From this center, engineers monitor the Internet for trends in spam and virus attacks and create updated definitions that are automatically downloaded by the Barracuda Spam Firewall.

Understanding Spam Scoring The Barracuda Spam Firewall (BSF) examines all the characteristics of a mail message and uses a complex system of scores to determine whether a message is spam. When an email reaches the spam scoring filter, the BSF assigns scores to all the properties of the message. Examples of what BSF examines are a message's header and subject line for offending characters or words, the percentage of HTML in the message, and whether a message contains an "unsubscribe" link.

The BSF is capable of Bayesian Learning, whereby it considers the characteristics of email that you have specified as being SPAM or NOT SPAM to compute scores. Finally, BSF uses email fingerprinting, the identification of exact or substantially similar spam messages via the Barracuda Central clearing house. All these methods help the BSF determine a message's spam score, which is displayed in the message's headers. The final score is compared to spam scoring limits to determine whether the message is considered spam and what is done with it after spam evaluation completes.

Spam Scoring Limits
Spam scoring limits are based on two settings:

Tag: Messages with a score above this threshold, but below the Block setting are delivered to your mailbox with the word [SPAM?] added to the subject line. Any message with a score below the tag setting is automatically allowed.

Block: Messages with a score above this setting are not delivered to the recipient, and a non-delivery receipt (NDR/bounce message) is sent to the sender from the BSF. Barracuda is 100% sure these messages are spam because the from address appears on the "200 Worst Spammers List".