The mailing list for listmasters using Sympa

[David Verdin <address@concealed>]

Hi Paul, sorry for this late answer,

This message means that it took Sympa 1358 seconds to evaluate if the 
message is allowed to be sent.

Basically, this is the time elapsed while Sympa successively performs 
the following operations (in this order):

- Display the message "Processing message for <list name> with priority 
<priority> , <message id>"
- Check message id to prevent loops
- Check that message is not a command faulty addressed to a list
- Check that the message is not a return receipt
- Check that the message is not too large
- If the list config defined topics, check the message tagging
- Check send scenario
- If the scenario allows this message sending, display the log "Message 
for <list name> from <sender address> accepted (<time elapsed> seconds, 
<numer of recipients> sessions), size=<message size> "

It's weird that it takes so much time for such simple operations.

Regards,

Paul Borghese a écrit :
> What does the following line represent?
>
> Feb 20 03:37:30 lists sympa[5631]: Message for ccielab from
> address@concealed accepted (1358 seconds, 54 sessions), size=4995
>
> Is it saying it took sympa 1,358 seconds to receive one 4k message?
>
> Paul
>
> -----Original Message-----
> From: Paul Borghese [mailto:address@concealed] 
> Sent: Wednesday, February 20, 2008 4:01 AM
> To: address@concealed
> Cc: address@concealed
> Subject: RE: [sympa-users] Queueing time of over a day
>
> Ok, I stopped sympa and moved the msg folder to msg-old and created a new
> msg folder.  I moved two messages from the old folder into msg and also
> posted two messages of my own.
>
> I then restarted the sympa processes.  The first message was processed
> immediately, but now 20+ minutes later the remaining four are still in
> queue.
>
> The only log message is the following:
>
> Feb 20 03:42:44 lists sympa[5907]: Processing message for ccielab with
> priority 5, <address@concealed>
>
> Feb 20 03:42:45 lists sympa[5907]: Cannot read
> /home/sympa/expl/ccielab/message.footer
>
>
> I realize there was an error with message.footer, but frankly that has been
> the case for the last five years.
>
> Is there any way to put sympa in a debug mode where I can see what is
> happening?
>
> Paul
> -----Original Message-----
> From: address@concealed [mailto:address@concealed] 
> Sent: Wednesday, February 20, 2008 3:14 AM
> To: Paul Borghese
> Cc: address@concealed
> Subject: Re: [sympa-users] Queueing time of over a day
>
> Paul Borghese wrote:
>   
> > For example, there is a message in /var/spool/sympa/msg that I sent hours
> > ago.  It is just sitting there waiting to be picked up by sympa.  In about
> >     
> a
>   
> > day or so, sympa will pick it out of the queue and deliver.  Sympa is
> > currently sending messages that were placed in the queue a day ago.
> >
> > Any suggestions on what I can try?
> >   
> >     
> How many mails are waiting in spool msg ? May be there is a lot of old 
> bad messages in the spool that make "readdir"  very long operation.  I 
> don't remember how this version deal with spools. I known it is much 
> more better now. Did you check logs ? I doubt that Sympa is waiting 
> without doing anything for hours. Logs will explain the problem.
>   
> > Also, do you know if I can upgrade directly to the most recent version of
> > Sympa?
> >     
> You should but you must check carefully the change log with a special 
> attention to lines related to incompatible changes (they are starting 
> with "****") . One of the major change is that now database is required 
> and parameter data_source in list config file must be set to 'include2'.
>
> Your problem is that automatic upgrade feature is newer then the versioj 
> you are using so it may be a problem when upgrading. Be aware that you 
> must  stop sendmail, apache ans all sympa process when upgrading in 
> order to avoid concurrent  access to datas that Sympa upgrade process is 
> changing.
>
> Of course for such upgrade, you have to  save your database and all 
> Sympa lists environment.
>   

--
David Verdin
Comité réseau des universités