At 10:41 6/19/97 CDT, you wrote:
|One of our lists experiences multiple deletes, but I haven't been able
|to figure out why...
|
|The latest round occurred yesterday.
|
|Checking the console logs I see that the message comes from LMAIL:
|>>Automatic non-delivery report processing for COMPMED:
|>>->Deleted [log in to unmask]
|>>->Deleted [log in to unmask]
|
|But nothing in the LMAIL log indicates why and what error occurred to
|cause the delete.  And while we had two addresses from MAIL.ARC.NASA.GOV
|there were also 7 addresses deleted from CCMAIL.BMS.COM.
|
|I also used a listserve distribution test...
|>> //Test JOB
|>> Distribute mail Ack=mail
|>> //To DD *
|
|to the deleted addresses and everything seemed to work out fine...at
|least I received a 'successful delivery' message.  So far, no reply
|from the addresses themselves.

The things to look for in most-to-least likely:

1) change of IP address (IBM's SMTP is notorious for not honoring TTL of
the MX/A records (assuming that DNS was appropriately reconfigured)

2) though your DISTRIBUTE would seem to make sense, it is not used in the
correct context.  To establish the context, REVIEW the COMPMED list, and
compose a LISTSERV DISTRIBUTE DEBUG=YES job and send THAT to the
LISTSERVing host.  Observe the output to determine the last site of the
LISTSERV Distribute Backbone for the boucning sites.  (BTW, not all
LISTSERV host sites use the same mail transport, thus some might generate
an SMTP HELO, while others an SMTP EHLO.  Some broken SMTP servers don't
support the latter).  If the last LISTSERV Distribute Backbone site is a VM
host, you might want to check to see if the MAILER has a proper configuation.

3) possibly a corrupted DNS cache used of the last L.D.B.
--
mailto:[log in to unmask]          Tel: +1 814 863 1843
31 Shields Bldg;  University Park,  PA   16802-1202 USA
Powered by: LISTSERV, Eudora, Netscape, mIRC, FreeAgent