Thoughts on Barracudas?

James R. Stevens jstevens at
Mon Aug 21 15:39:59 IST 2006

Quote of the day:  "Bayes is like a fast woman, keep an eye on it and be
prepared for suprises, and it is worth having. Just don't turn your back
on it."

Super Dave, your alright in my book!! 

-----Original Message-----
From: mailscanner-bounces at
[mailto:mailscanner-bounces at] On Behalf Of DAve
Sent: Monday, August 21, 2006 8:26 AM
To: MailScanner discussion
Subject: Re: Thoughts on Barracudas?

Julian Field wrote:
> Hash: SHA1
> Folks,
> I have had some comments from a few people leaving the mailing list as

> they are ditching their MailScanner setups and switching to Barracuda 
> applicances instead. They claim that things worked fine when they 
> first installed MailScanner, but gradually more and more spam is 
> leaking through, to the point where they have decided to abandon it.
> Here is what he said:
> "I've tried Mailscanner on FreeBSD for almost one year. It worked 
> great for about two months, then after every upgrade it began to let 
> more and more spam through. I've tried everything to fix it and just 
> got tired of my users complaining of increased spam.
> "It wasn't worth the headache. Your forums indicate that there are 
> numerous people experienced the same problems I have encountered.
> "I have since purchased a Barracuda SPAM 200 firewall. This device has

> worked much better."
> What is your opinion on the Barracuda appliance?

Not good overall, I deal with a *lot* of shrinkwrap administrators who
have no understanding at all of how to use the app, how it works, or
what each option really does.

> How easy is it to use?

Very easy, which is why so many get purchased. Just "plug it in" and
watch your spam disappear.

> Does it actually work?

I believe it can work in the hands of a capable administrator. In most
installations I believe it works "Okay", see below for an explanation.

> Can it survive the loads they say it can?


> And, of course, how does it compare with MailScanner?

See below.

> Please be open and honest, and as impartial as you can.

My experience comes from two places. Clients who have a Barracuda and do
not use our service (MailScanner based) and clients who use our service
in front of, or in lieu of a Barracuda. We are an ISP with a wide
variety of mail traffic. Corporate (lots of word docs and leek speak
believe it or not), Worldwide (Pacific Rim, Eastern and Western European
languages), marketing (everything SA is designed to catch ;^), technical
(large attachments), dialup users (it's not spam it's magic lotto
numbers, that whitelisted Ebay monthly report is spam! Why do I keep
getting it?).

Based on those scenarios, my experience is that the Barracuda can be
configured to be very effective for a single installation such as a
small corporate LAN. However, you must be willing to accept it's

If you purchase a product chosen from a shiny four color add with pretty
people looking pensive in front of a massive wall of servers, you
believe the product works and that the product supplier has designed it
well. You see the limitations and you accept them. You make your
argument to the board, you install the product, it works well enough,
the company feels you are 'doing something' about the problem. This is
where the Barracuda succeeds.

If you purchase a product from your Service Provider, the SP will
invariably need to offer some type of advantage over doing your own
filtering. Many of the things I have been asked to do for clients would
never have been possible with an OTS system. The fine grained control
offered by MailScanner far exceeds anything Barracuda could possibly
offer. The advantage offered is "configurability" This is where
MailScanner succeeds.

Does that help?


PS. If a MailScanner box is failing to catch spam over a period of time
I would think the answer was right in front of the user. Bayes. I had
never been a fan of Bayes as it seemed to always cause more load and
headache than it was worth. However, out of sheer need, I have
configured and run Bayes for the last three weeks and I have to say it
works much much better in 3.1.X than it did in 2.[4-6].X. But I can see
where a combination of Bayes, AutoLearning, and lack of attention could
lead to very bad capture rates.

Bayes is like a fast woman, keep an eye on it and be prepared for
suprises, and it is worth having. Just don't turn your back on it.  ;^)

Three years now I've asked Google why they don't have a logo change for
Memorial Day. Why do they choose to do logos for other non-international
holidays, but nothing for Veterans?

Maybe they forgot who made that choice possible.
MailScanner mailing list
mailscanner at

Before posting, read

Support MailScanner development - buy the book off the website! 

This message has been scanned for viruses and dangerous content by
Athens Hyperion Scanner, and is believed to be clean.

This message has been scanned for viruses and
dangerous content by Athens Hyperion Scanner, and is
believed to be clean.

More information about the MailScanner mailing list