The mailing list for listmasters using Sympa

[David Verdin <address@concealed>]

Hi,

Would it be possible that a lot of people have their reception 
preferences set to "no mail"?
If Sympa crashed while processing this message, you should receive a 
crash notification the next time you launch Sympa. Additionnally, you 
would not be able to send any message thereafter, the Sympa processes 
being stopped.
If go on with the hypothesis of a Sympa failure, something should 
trigger it. Is it always the same address that is the last to be processed ?
What do the Sympa logs say after having processed the message? Do they 
sop or do they continue as if nothing happened?
What version of Sympa do you use?
If Sympa crashes, it dumps its last error messages in a file called 
####.stderr (where #### is the id of the process that just crashed) 
located in the tmp spool. You could find useful informations there.

Regards,

Gregor Hoffleit a écrit :
> We're testing a setup with a huge list (130.000 subscribers).
> When I send a message to the list, Sympa pumps out 1400 mails to the
> Postfix MTA, covering 28.000 addresses. Then it simply stops delivering.
>
> Strange thing: It leaves a message in the syslog that sounds like a
> success message:
>
>     sympa(message)[11853]: main::DoMessage() Message for test-list from address@concealed 
> accepted (466 seconds, 1439 sessions, 100243 subscribers), message-id 
> <address@concealed>, size=43874
>
> But, as you can see, something's wrong here: 1439 mails with at max 25
> addresses is much less than the expected 100243 recipients. And in fact,
> if I look into the mail log, there are less than 29000 recipients.
>
> And it reproducably stops at 1439 sessions, when I rerun the message.
>
>
> I would expect that Sympa would log some error message or at least a
> warning, IF if would drop some messages deliberately.
>
> Is it possible that Sympa just crashes here silently?
>
>
> Any hint where I could go and look for more information,
>
>
> Thanks a lot in advance,
>     Gregor
>   

--
David Verdin
Comité réseau des universités