I don't understand how the spoolthreads are able to read new messages
from the spool but not to write new ones!
They should get connection problems while reading too!?
BTW, I currently solved this scenario by implementing my own mailing
list mailet (it's not a mailing list but a newsletter with mailmerging
features) that prepare all the generated mails in a temporary spool and
then a second mailet takes all the mails from the temporary spool and
send them.
I did this because I want to be sure that the newsletter start only when
the full merging has been possible.
My code is not generic enough or generally ready to be committed and
currently contains too much "private" code to bu published but I have
plans to publish it in future (read as months/year!?).
The "Store" component let you to lookup for custom repositories (create
new repositories on the fly).
- The "template message" enter the spool
- A first mailet assign a new Attribute to the message to say what
repository should be used
- ToProcessor move the message to another processor (the message is
updated on db only on processor change)
- A mailet start merging the results in the given processor and add a
mail attribute to say it finished
- ToProcessor move the message to a third processor
- A mailet start moving the messages from the temporary repository to
the transport repository.
PS: I never used the list managers included in James because I need
mailmerging.
Hope this help.
Stefano
JWM wrote:
I really appreciate all the time you've spent explaining/assisting me in
this.
What is happening in my scenario is that it the maillist gets part way into
sending emails to the list members, then the failure occurs and it aborts.
But it immediately restarts and starts sending from the top of the list
again. It fails again part way through the list, and restarts. If the mail
went out at night or when I can't be contacted, this just keeps repeating
forever until somebody realizes it's gone crazy and notifies me, and can get
to a place to log in to the server and stop it. Once it starts, it will not
recover on its own. It appears that some random situation first puts it
over the edge, but continually restarting the maillist ends up stressing the
connection pool enough to make it feed on itself and never recover. When I
stopped the server, I had about 1500 emails in the spool table (the original
source email + all the individual ones I had sent). I just wiped these out
of the db to recover.
The recipient at the top of the list gets flooded. The last recipient in
the list likely never got the email at all.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
