Forum OpenACS Q&A: e-mail notifications

Collapse
Posted by Tom Mizukami on
I have received this e-mail notification of a bboard posting 133
times.

Woopie!, Easy! That doesn't hurt at all.
THANK YOU Michael.

-Bob

I can remember another instance where I received a posting multiple
times. Does anyone know what causes this, the poster, OpenACS, or our
mail server configuration. It's not a big deal I was just wondering.

Collapse
Posted by Don Baccus on
That's very bizarre.  The forum software calls ns_sendmail once for each e-mail address, and weeds out duplicates beforehand.

ns_sendmail opens a socket, builds an SMTP message, ships it off to  the addressees, and closes the socket.  The forum software does an individual ns_sendmail for each person's alert message, which is why the names pop ... up ... one ... at ... a ... time ... slowly on the post-confirmation page shown to the user.

I've never had anything like this happen on my server - I use my ISP's  SMTP server.  I've never received bombardment like this from any ACS forum I've asked for alerts from, and that's quite a few forums both OpenACS and ACS Classic.

I vote for your e-mail server configuration unless others confirm they've had the same problems.

Collapse
Posted by Bob OConnor on

    Woopie!, Easy! That doesn't hurt at all. THANK YOU Michael.

I was the one who wrote the answer once, that you got 133 times. Maybe I was irrationally exuberant!!! 😊

So I thought I'd respond here. I doubt that it's me. I get all the posts from OpenACS directed to a folder in my email program and a quick check shows no duplicates...

-Bob

Collapse
Posted by Edmund Lian on
I've seen this behavior before with other apps. Getting multiple messages can happen when the connection between the orginating MTA and receiving MTA is broken after the originating MTA completes message delivery and is disconnecting, but before the receiving MTA acks the disconnection request. When this happens, the originating MTA thinks the message wasn't delivered, and tries again.

One would think this kind of failure would be covered by the SMTP protocols. OTOH, it might be the specific implementation of the originating MTA involved. Never bothered to dig down further than this since the problem went away when a more reliable link was available.