LSTSRV-L Archives

LISTSERV Site Administrators' Forum

LSTSRV-L

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Topic: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Valdis Kletnieks <[log in to unmask]>
Tue, 16 Feb 1999 14:17:51 -0500
text/plain (2100 bytes) , application/pgp-signature (291 bytes)
On Tue, 16 Feb 1999 13:26:31 EST, Stanley Horwitz said:
> Can I generate a probe of each subscription on my list regardless of the
> age of the subscription and have any subscription that elicits a rejection
> of the single probe be removed from the list as soon as the rejected probe
> message reaches Listserv? If so, how? This is with Listserv 1.8d. I have
> read through the docs for 1.8c and 1.8d, but I cannot seem to get the
> probe to go out immediately for all subscriptions. Any tips will be
> appreciated.

I'd suggest being careful on removing people after a *single* failed
probe.  The *last* thing you want to do is nuke people off their list
just because their server had a problem that was corrected soon
thereafter (full /var/spool space, a corrupted database, system out
of resources due to some *other* user having a program in a loop,
etc etc ad nauseum).

And yes, there are many systems out there that will return a '5xx'
code for some of these conditions, because it's something that the
MTA doesn't *know* for a fact will clear up in a few minutes. Appendix E
of RFC821 is pretty specific about it:

            5yz   Permanent Negative Completion reply

               The command was not accepted and the requested action did
               not occur.  The sender-SMTP is discouraged from repeating
               the exact request (in the same sequence).  Even some
               "permanent" error conditions can be corrected, so the
               human user may want to direct the sender-SMTP to
               reinitiate the command sequence by direct action at some
               point in the future (e.g., after the spelling has been
               changed, or the user has altered the account status).

In particular, if the user's mailbox is too large, the problem will
clear itself up as soon as he reads his mail.  Seems pretty harsh
to unsubscribe him just because he was out sick with the flu and
didn't read his mail for a few days... ;)


--
                                Valdis Kletnieks
                                Computer Systems Senior Engineer
                                Virginia Tech



ATOM RSS1 RSS2