Loading ...
Sorry, an error occurred while loading the content.

Lost connection while sending end of data

Expand Messages
  • Diego Alencar Alves de Lima
    Hello all, I currently have a postfix MTA working together with a content filtering solution (Trendmicro s IMSS)and I ve been receiving the following errors
    Message 1 of 4 , Aug 31 8:13 PM
    • 0 Attachment
      Hello all,

      I currently have a postfix MTA working together with a content filtering
      solution (Trendmicro's IMSS)and I've been receiving the following errors
      constantly:

      (lost connection with localhost[127.0.0.1] while sending end of data --
      message may be sent more than once)

      The message are never delivered and as of now it's causing a queue of
      over 100 messages and filling up slowly. Could this be a problem with
      postfix?

      I've read some suggestions that this error could be caused by path
      discovery, but since both postfix and the content filter are in the same
      machine I don't think it could be this.

      How can I further debug the problem? It's a problem that is really
      becoming critical. Thank you.

      --
      Diego Lima
      www.diegolima.org
    • Wietse Venema
      ... The SMTP server on 127.0.0.1 did not respond to end of data within $smtp_data_done_timeout seconds (default 600s, or 10 minutes, as recommended by the
      Message 2 of 4 , Aug 31 8:30 PM
      • 0 Attachment
        Diego Alencar Alves de Lima:
        > Hello all,
        >
        > I currently have a postfix MTA working together with a content filtering
        > solution (Trendmicro's IMSS)and I've been receiving the following errors
        > constantly:
        >
        > (lost connection with localhost[127.0.0.1] while sending end of data --
        > message may be sent more than once)

        The SMTP server on 127.0.0.1 did not respond to "end of data" within
        $smtp_data_done_timeout seconds (default 600s, or 10 minutes, as
        recommended by the RFC).

        > The message are never delivered and as of now it's causing a queue of
        > over 100 messages and filling up slowly. Could this be a problem with
        > postfix?

        The SMTP server on 127.0.0.1 does not respond to "end-of-data". Is that
        a problem in Postfix?

        > I've read some suggestions that this error could be caused by path
        > discovery, but since both postfix and the content filter are in the same
        > machine I don't think it could be this.
        >
        > How can I further debug the problem? It's a problem that is really
        > becoming critical. Thank you.

        Find out why the server on 127.0.0.1 does not respond to "end-of-data".
        If "end-of-data" does not make sense to you, find someone who
        understands SMTP. Use tcpdump, strace, and other diagnostic tools
        to find out what it's doing instead or responding to "end-of-data".
        If this does not make sense to you, find someone who is familiar
        with diagnostic tools such as tcpdump and strace, and who understands
        what their outputs mean.

        Wietse
      • Diego Alencar Alves de Lima
        ... The messages do not take long to be deferred. I see the error as soon as I flush or requeue these particular messages. And this doesn t happen to all
        Message 3 of 4 , Aug 31 9:55 PM
        • 0 Attachment
          Em Sex, 2007-08-31 às 23:30 -0400, Wietse Venema escreveu:
          > The SMTP server on 127.0.0.1 did not respond to "end of data" within
          > $smtp_data_done_timeout seconds (default 600s, or 10 minutes, as
          > recommended by the RFC).

          The messages do not take long to be deferred. I see the error as soon as
          I flush or requeue these particular messages. And this doesn't happen to
          all messages, just some specific few.

          Could there be another reason for this message other than a timeout
          while waiting for the smtp response?

          Thank you.

          --
          Diego Lima
          www.diegolima.org
        • Wietse Venema
          ... I was not aware that the drop was instantaneous. ... Apparently, the filter aborts quickly. The reason for this is likely written to a logfile, provided
          Message 4 of 4 , Sep 1, 2007
          • 0 Attachment
            Diego Alencar Alves de Lima:
            > Em Sex, 2007-08-31 ?s 23:30 -0400, Wietse Venema escreveu:
            > > The SMTP server on 127.0.0.1 did not respond to "end of data" within
            > > $smtp_data_done_timeout seconds (default 600s, or 10 minutes, as
            > > recommended by the RFC).
            >
            > The messages do not take long to be deferred. I see the error as soon as
            > I flush or requeue these particular messages. And this doesn't happen to
            > all messages, just some specific few.

            I was not aware that the drop was instantaneous.

            > Could there be another reason for this message other than a timeout
            > while waiting for the smtp response?

            Apparently, the filter aborts quickly. The reason for this is likely
            written to a logfile, provided that the file permissions aren't
            broken to the point that the filter can't write its log.

            If nothing is logged, use strace, truss, etc. to diagnose the process.

            Wietse
          Your message has been successfully submitted and would be delivered to recipients shortly.