sendmail sometimes refuses connections.. anyone have simularproblem

[Matt Doherty]

Ok Thanks for the tip!
Yes the webmail program is on the same machine. And webmail was sending mail
fine at the time sendmail refused mail sending from email client programs...
>The next time it happens to you, try to list the listening sockets, and
t>he processes ... I'm sure it will help you to diagnose the problem! :)

You mean run the "top" command i suppose. Right?
But how and why would sendmail refuse to listen on the port. DOS attacks
wern't present at the time either. Nothing suspicious was happiening during
the time of failure. cpu was at 7% 1GB of memory had 25% free, maillog was
calm (tail -f maillog).. Problem is, to be able to catch it, I must be here
in the wee morning hours with toothpicks holding my eye lids open.. seems to
occur anywhere between 11pm - 7am.. I thought it could be a cron job
affecting this, but there isnt any problems of any runaway processes going
on either..

Anyways, Thank You for the reply!

In a world without walls or fences, who needs Windows and Gates?

  -----Original Message-----
  From: Marc Perea [mailto:marc.perea at ELECTRONIC-GROUP.COM]
  Sent: Friday, July 12, 2002 12:00 PM
  To: MAILSCANNER at JISCMAIL.AC.UK
  Subject: Re: sendmail sometimes refuses connections.. anyone have
simularproblem


  On Fri, 12 Jul 2002 10:35:50 -0400
  Matt Doherty <Matthew_doherty at DATAWATCH.COM> wrote:

  >
  > I know mailscanner runs a script every hour via cron to check its pid..
  > I'm wondering if it sometimes has difficulty starting the sendmail
  > daemon if it had to..

  My understood is that mailscanner doesn't do anything with sendmail every
  hour. It only restarts himself to prevent memory leaks. (Correct me if I'm
  wrong)

  > I'm not quite sure why this happens. I have read my maillogs, xfer logs,
  > kernal messages, and find nothing to conclude this issue. This morning I
  > started to receive calls that sendmail was down. Our web based email
  > could send mail but client email programs were being refused..
  > Obviousley that ment sendmail was running fine right? I restarted
  > Mailscanner for the hell of it, and now both web based mail and client
  > email clients can send.. So its gotta be sendmail right? or maybe it had
  > some effect on the xinitd? Anyone here know this problem? This has been
  > happening on and off ever since I installed mailscanner. Though it could
  > be a coincidence that it started after the install, I'm not sure.
  >

  That probably means that your sendmail stopped to be in background
  listening for connections in port 25, that's why the client email programs
  cannot remotely access to it, but the webmail program (that I suppose it
  was running on the same machine) was still able to send emails directly
  invoking /usr/sbin/sendmail in foreground mode.

  The next time it happens to you, try to list the listening sockets, and
  the processes ... I'm sure it will help you to diagnose the problem! :)

  Hope that helps you.

  Cheers,

  --
  Marc Perea - System Administration Staff
  Mail: marc.perea at electronic-group.com
  Tel: (+34) 93 600 23 23
  Fax: (+34) 93 600 23 10
  ----------------
  Electronic Group - http://www.electronic-group.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mailscanner.info/pipermail/mailscanner/attachments/20020712/783ac900/attachment.html