On Thu, 07 Sep 2000 11:40:21 EDT, Kathy Montgomery <[log in to unmask]>  said:
> We've been having this problem, too, with "wa" crawling (and LISTSERV overall)
> when one of our 1000+ subscriber lists gets a posting.   I just read in this
> thread that the SMTP_FORWARD_n variable is the solution.  Could someone clarify
> its use for me, with some simple examples?  I'm having trouble understanding the

I run with 'SMTP_FORWARD_1=listserv.vt.edu' (i.e. itself) and it works just
fine.  The crucial point here for slow 'wa' response time is that then the
main lsv process can farm out the time-consuming part of handing stuff to
the local sendmail and come back to process the next request as soon as
the child process has *started* to hand the mail to Sendmail, rather than
having to wait till it has ALL been passed to Sendmail.
--
                                Valdis Kletnieks
                                Operating Systems Analyst
                                Virginia Tech