RE: [Declude.JunkMail] Question about load balancers and source IP

[Landon Jenkins]

 Does Declude/IMAIL care about the IP address that's making the
 connection?
 In other words, does it use that IP address for its tests? If
 so, will HOP=1 fix this?

I have never used a Foundry Load Balancer so my response may be way off. I
am assuming it is not functioning as a MTA, but is simply rewriting the
source IP portion of packets.

If the source IP of incoming packets destined to your declude gateways is
replaced with that of your load balancer, it does not constitute an
additional hop as far as declude is concerned. The message header will
likely report the connecting IP address (in this case, being that of your
load balancer if doing source NAT) but also with the sending MTA's HELO.

The IP tests including RBL lookups and SPF checks are based on the last
reported sending MTA's IP address. Also, other DNS tests such as REVDNS
won't mean much of anything anymore since declude will only be checking for
a PTR for the IP of your load balancer. Non-IP based tests should work fine
I would think.

As much of a pain as it might be, I would stick with DSR if you don't want
to lose declude's IP-based tests.

Landon



---
This E-mail came from the Declude.JunkMail mailing list.  To
unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
type unsubscribe Declude.JunkMail.  The archives can be found
at http://www.mail-archive.com.

[Declude.JunkMail] Queue freezing, delivery issues

[Landon Jenkins]

I have seen some recent talk about queue problems and while my issues may be
unrelated, it is definitely a problem.

I am running IMail 7.15 and Declude 1.82 Junk/Virus Pro.

Starting Tuesday, June 14, at random times, IMail will accept incoming SMTP
connections, accept the mail, and place it in the queue. However, delivery
is not immediate and in fact, the files in the queue never process. The
files in the queue quickly accumulate and local and remote delivery halts.

When I restart the SMTP service, new incoming messages are processed and
delivered in real time, however the files in the queue must be processed
manually. Once I start the manual queue processing, the queue quickly
empties and all is well until the next time it happens. This is happening at
least once per day. I have not seen any trend in the queue or logs that
suggest what the problem might be.

My configuration has not changed for several months. I have checked the
resolution and responses for RBL's that I have configured in global.cfg. All
looks well there.

I do notice that a similar thread regarding this issue has been discussed at
the IPSwitch forum: 

http://www.ipswitch.com/forums/shwmessage.aspx?ForumID=10MessageID=1260Top
icPage=1

My system config:
Windows 2000 Server w/latest patches
512Mb RAM
Spool is on a pair of mirrored SCSI drives

I appreciate everyone's ideas.

Landon

---
This E-mail came from the Declude.JunkMail mailing list.  To
unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
type unsubscribe Declude.JunkMail.  The archives can be found
at http://www.mail-archive.com.