Re: Large Queue Of Tempfails

On 1 Apr 2010, at 19:45, Davide Libenzi wrote:
On Thu, 1 Apr 2010, Sabahattin Gucukoglu wrote:
>> At any given time I have about 100 mails destined to go nowhere due to 
>> forged mail setting off a challenge, because the MX record is just "." 
>> or sometimes "dev.null".  It would be nice if such errors were detected 
>> immediately.  When these mails are in the queue, any new mail arriving 
>> by SMTP is delayed quite noticeably.  Is there anything I can do about 
>> this except
>> find /var/xmail/MailRoot/spool ! -type d -delete
>> from time to time?  Would it be possible to deal with new mail received 
>> by SMTP first, then the stuff in rsnd directories?
> 
> The command above would be a really bad idea, as it'd nuke the spool :)
> On top of that, if you do that when XMail is running, you are going to 
> mess up with it, since you are removing content from within its domain.
> At the moment you'd need to do it externally, by parsing the spool.
> But if you want to remove stuff from it, you better stop XMail before, and 
> clean all the associated files inside the spool.

Yep, XMail stopped by first touching MailRoot/.shutdown and waiting for the file to disappear for a few seconds.  Then the above command, after checking that the files are only slog/* and rsnd/*, i.e., mails being retransmitted which I know I didn't send.  The directories are left alone.  Then restart, and it's snappy again.  I will look at increasing the number of queue threads if it keeps happening.

Cheers,
Sabahattin

_______________________________________________
xmail mailing list
[email protected]
http://xmailserver.org/mailman/listinfo/xmail