I believe that what you are getting at is a very likely explanation for this. I had never seen this before the other day, and it seems that most of the reports are coming from IMail 8 users who are using IMail to run tests before Declude gets the message. I figure that my system's window is only about 1/5 of a second, but if IMail is doing address verification, this could be as much as a minute (based on previous discussions). That would make this 300 times more frequent on your system than it would be on mine with other things being equal.
The good news I guess is that if it is skipping IMail's tests also, and the delays in processing some things like address verification are as long as indicated, IMail will want to see fit to fix this quickly, at least in version 8. I really don't want to upgrade, so lets hope for my sake and many others that they fix it in version 7 as well. This is a major flaw in their process.
I'm guessing that the totality of your IMail filtering might be something to look at instead of just that one test. I would imagine that it could take up to 2 seconds for all of the DNS-based tests to be run. The fact that you see so many of these and I have only seen one suggests strongly that the delay itself with IMail 8 is the reason. I certainly would have caught this many other times in my own inbox considering the volume that gets sent to it.
Matt
Kami Razvan wrote:
Hi;
Some common observations about Declude not seeing emails:
- Last night I personally had 2 emails that were not seen by Declude and 4 were in the Admin mailbox. At least these are the ones that I have seen. Of all of these none have the IMail spam headers.
- In our configuration we do all the IP4r tests in IMail and add header for Declude to analyze. It is as if IMail never added the headers.. Since none are there. Could it be that IMail somehow skips its own spam test? Should we not expect if IMail has done all that it was to do the headers from its spam test would show up? Why are they absent?
- Our system is also setup to delete emails that fail 13+ tests by IMail and not even bother Declude with it.
- Matt: We are doing address verification. I will disable it and see if that can help. I have found this to be a very valid test and I know it is resource intensive. I will disable it.
- Declude's action for all these emails should have been delete. Based on the conditions that are present in all of these emails none should have made it since they have email addresses in the CC field that are listed in our delete filter if found in the recipients.
May be if we try to present the common factors of the ones that are being skipped we can find what is causing it. I know for sure we are getting 5+ a day..
Regards, Kami
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Matthew Bramble Sent: Sunday, December 07, 2003 7:37 PM To: [EMAIL PROTECTED] Subject: Re: [Declude.JunkMail] Declude not taking action
Kami Razvan wrote:
I have spent a lot of my years in the field of mathematics. A study done a while back and it is related to data-mining stated.. "men buy baby diapers and orange juice on Tuesdays more than any other day of theweek".
Sure it's useful, what it says is that there is something else going on here at least on your system. It could be possible that this can happen during the entire duration of processing the queue instead of the instant where it is initiated. It could be related to using some of IMail 8's filters before handing off to Declude, for instance the address verification which can produce delays of many seconds and make your window much wider. And of course the frequency of running your spool and your processor load can affect this, though to a smaller degree than your experience suggests.
Matt
--- [This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)]
--- 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.
