OSEC

Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com
 
From: Alastair Sherringham (alastaircalliope.demon.co.uk)
Date: Tue Feb 26 2002 - 16:59:10 CST

  • Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

    Hello,

    I had a problem today with a postfix process that I hope someone can
    help with. Nothing critical, and probably not the fault of postfix
    itself, but something I'd like to understand more.

    On this particular machine I am running Postfix-20010228-pl03.

    This afternoon I noticed some mail stuck in the active queue via mailq
    i.e.

    [rootrum alastair]# mailq
    -Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient-------
    AEF077D026* 3058 Tue Feb 26 12:07:02 richardnucoda.com
                                             xxxxxbbc.co.uk

    (the xxxx is a real email address I see no reason to post)

    I noticed this at 5pm.

    Logs ;

    Feb 26 12:07:02 rum postfix/smtpd[692]: connect from pernod[192.168.0.109]
    Feb 26 12:07:02 rum postfix/smtpd[692]: AEF077D026: client=pernod[192.168.0.109]
    Feb 26 12:07:02 rum postfix/cleanup[693]: AEF077D026: message-id=<LEEHIOIAIDFGAIHOKBGBCEAGCPAA.richardnucoda.com>
    Feb 26 12:07:02 rum postfix/qmgr[30563]: AEF077D026: from=<richardnucoda.com>, size=3058, nrcpt=1 (queue active)
    Feb 26 12:07:02 rum postfix/smtpd[692]: disconnect from pernod[192.168.0.109]

    I could not get this mail to go with a 'postfix flush' (which is ok I
    guess since 'flush' is for deferred retries I think).

    Other mails, to the same domain have been sent without problem all
    afternoon.

    Trying to figure this out I looked at network connections (netstat) and
    then, using 'ps' ;

    [rootrum alastair]# ps -o pid,tt,user,cmd,tmout,flags,time,cputime,pri,wchan -p 695
    PID TT USER CMD TMOUT F TIME TIME PRI WCHAN
    695 ? postfix smtp -t unix -u - 100 00:00:00 00:00:00 39 tcp_recvmsg

    This smtp process appeared 'stuck' waiting on 'tcp_recvmsg'.

    Restarting postfix cleared this and the message went.

    Why was it stuck? What might cause this? Would it ever have been
    deferred or cleared?

    Would a 'kill' on this process itself allowed delivery?

    Any help in understanding this would be welcome.

    Many Thanks,

    -- 
    AS                                  | 
    alastaircalliope.demon.co.uk       |
    http://www.calliope.demon.co.uk     |            PGP Key : A9DE69F8
    -------------------------------------------------------------------
    -
    To unsubscribe, send mail to majordomopostfix.org with content
    (not subject): unsubscribe postfix-users