[Mimedefang] I still have hair left
Tim Pushor
timp at crossthread.com
Tue Jul 22 11:36:02 EDT 2003
Mitch,
Yeah, I don't think its a timeout, since I get the 'Please try again
later' error in OE immediately after clicking send. I have also disabled
the antivirus checks and am only checking for bad attachments (just to
get something running).
I did get a little further last night.
MIMEDefang *is* running. When I add the -k flag to the startup script,
and send an email - it creates the temporary directory with the files in
it that I later run mimedefang.pl on. This proves that sendmail and
MIMEDefang are actully communicating. As soon as I remove the -k flag is
when it doesn't seem to work properly.
I verified that mimedefang.pl does in fact run. I added a line at the
beginning that just 'touches' a file to verify that it is running. It is.
So:
- So far, adding the -k switch shows that mimedefang is taking the
message apart and setting the stage for mimedefang.pl
- Running mimedefang.pl on this directory does what I expect (and logs it)
- without the -k switch, I have proven that mimedefang.pl does in fact
run, but gives me the 'Try again later' error immediately and nothing
gets logged
I guess I can try to truss a sendmail process and see whats happening
over the socket. Or I could just get a rope .... ;-)
Thanks again,
Tim
Mitch at 0Bits.COM wrote:
>Hi Tim,
>
>Your sendmail is error is quite strange. Look at the delay= line
>in yours
>
>
>
>>Jul 21 22:45:14 calgw sendmail[5099]: h6M4jDPc005099: to=<timp at crossthread.com>, delay=00:00:01, pri=30425, stat=Please try again later
>>
>>
>
>and what i was getting before i fixed my syslog
>
> Jul 9 00:20:14 core sm-mta[28247]: h68NID1I028247: to=<mitch at 0bits.com>, delay=00:02:01, pri=140375, stat=Please try again later
>
>In mine, it is clearly timing out after 2 minutes odd, whereas your's
>seems to be effectively generating the error instantaneously (1s) like
>as you say, it's not finding mimedefang.pl at all to execute..
>
>Are you SURE you've got the multiplexor and the mimedefang.pl installed
>correctly in the right path for the multiplexor to spawn it ?
>
> % strings /usr/bin/mimedefang-multiplexor|grep mimedefang.pl
> /usr/bin/mimedefang.pl
>
>Are you sure your socket is set up correctly and is readable by
>the defang user ?
>
>If the answers to all above are yes, then the last thing to do is
>use your UNIX tools - in Linux's case "strace" and in FreeBSD probably
>the "truss" command to truss sendmail (using the -f flag to follow it's
>children) and see what it is failing on.
>
>Mitch
>
>
[snip]
More information about the MIMEDefang
mailing list