On 23/02/2007, at 10:18 AM, Dave Watkins wrote:

> Hi All
>
>
>
> I’ve just stumbled upon ASSP completely by accident and am hoping  
> to be able to use it to replace our spamassassin system but have  
> some basic questions that aren’t immediately obvious in the wiki or  
> the docs (unless I’m going blind J)
>
>
>
> Currently we have our incoming mail hitting Mailmarshal first,  
> which does virus and rules checking, validates the mail is to a  
> real user (during SMTP) and some RBL checks, then it goes to  
> spamassassin and gets marked in the subject if it’s spam and then  
> passed to Exchange. Outbound goes from Exchange straight out to  
> Mailmarshal then sent.
>
>
>
> Now the right way to setup ASSP seems to be Exchange -> ASSP ->  
> Mailmarshal -> Internet.
>
>
>
> Inbound seems to be Internet ->  ASSP -> Mailmarshal -> Exchange.
>
>
>
> All this is fine and setup shouldn’t be a problem, I can move RBL  
> checks away from Mailmarshal and have them running on ASSP, virus  
> scanning will stay on Mailmarshal as it has the processing  
> capability to handle it but my questions are as follows.
>
>
>
> Will I need to setup LDAP on ASSP to validate that incoming mail is  
> addressed to a valid user? I want to block these at SMTP to save  
> bandwidth, there seems to be some indication in the docs that ASSP  
> would initiate an SMTP connection to Mailmarshal to check if the  
> recipient is valid before it responds to the sending SMTP server  
> with an OK but I’m unsure if this is the case or not. Basically how  
> does ASSP validate incoming mail is going to a real user?

You can use a flat file (export it from Mailmarshal) or LDAP.
>
>
> At least in the initial stages I don’t want to change the behaviour  
> of our current system so I would prefer if ASSP didn’t block  
> messages but instead just marked them in the subject. Most of our  
> users have rules to move marked messages to a separate folder and,  
> at least in the short term I’d like this to continue. Once we have  
> confidence in our Bayesian DB I’ll switch it over to blocking by  
> default. It looks like the only way to accomplish this is by  
> turning on testing mode. Is that correct?

Yes, you can have everything in test mode, then turn off test modes  
one at a time, as you get comfortable with them.
>
>
> Is there any way to pre-populate the “Delaying” database during the  
> testing phase without actually having it have delaying turned on?  
> Ideally I’d like to be able to have the DB already populated with  
> IP’s of servers that haven’t sent us spam when I enable it so they  
> don’t receive delays.
Whitelisted addresses are not delayed - you can fill up the whitelist.

>
>
> Is there any scoring attached to failing an RBL check? RBL blocking  
> has given us some issues in the past when the major ISP’s in the  
> country manage to get themselves on an RBL list. All I can see is a  
> setting that defines how many different RBL checks have to fail  
> before the message is blocked, which admittedly may be enough in  
> our case but may also let in more spam. Is setting that to 2 (or  
> maybe 3) a good idea?

You can use Penalty Box to score things. Then once the threshold is  
exceeded, the sender is blocked.
>
>
> I think that’s it for now
>
>
>
> Thanks in advance

Good luck.

James.
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Assp-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-user

Reply via email to