[Mimedefang] mimedefang w/ multiplexor vs timeout error (update)

[Xiaoyan Ma]

Here is the update on my timeout error.

I enabled debuging last night and got more detailed error message related 
to timeout:

Apr  9 03:39:27 bear-eth0 mimedefang[23208]: [ID 893429 mail.debug] 
Entering envfrom (line 500)
Apr  9 03:39:27 bear-eth0 mimedefang[23208]: [ID 702342 mail.warning] 
h39AdQbS005474: Could not create directory /var/spool/MIMED
efang/mdefang-3E93F85F-0: File exists
Apr  9 03:39:27 bear-eth0 mimedefang[23208]: [ID 893429 mail.debug] 
Entering mfclose (line 1318)
Apr  9 03:39:27 bear-eth0 mimedefang[23208]: [ID 893429 mail.debug] 
Entering cleanup (line 1369)
Apr  9 03:39:27 bear-eth0 mimedefang[23208]: [ID 402246 mail.info] Exiting 
cleanup (line 1437) ret=SMFIS_CONTINUE
Apr  9 03:39:27 bear-eth0 mimedefang[23208]: [ID 521097 mail.debug] Exiting 
mfclose (line 1334) ret=SMFIS_CONTINUE
Apr  9 03:39:27 bear-eth0 mimedefang[23208]: [ID 521097 mail.debug] Exiting 
envfrom (line 581) ret=SMFIS_TEMPFAIL
Apr  9 03:39:27 bear-eth0 sm-mta[5474]: [ID 801593 mail.info] 
h39AdQbS005474: milter=mimedefang, action=mail, tempfail
Apr  9 03:39:27 bear-eth0 sm-mta[5474]: [ID 801593 mail.info] 
h39AdQbS005474: Milter: from=<daily_headlines at ms1.lga2.nytimes.com>
, reject=451 4.7.1 Please try again later
Apr  9 03:39:27 bear-eth0 sm-mta[5474]: [ID 801593 mail.info] 
h39AdQbS005474: --- 451 4.7.1 Please try again later
Apr  9 03:39:27 bear-eth0 sm-mta[5474]: [ID 801593 mail.info] 
h39AdQbS005474: <-- RSET
Apr  9 03:39:27 bear-eth0 sm-mta[5474]: [ID 801593 mail.info] 
h39AdQbS005474: --- 250 2.0.0 Reset state

I searched the archive and found David's reply a year ago:

*********************************************
On Mon, 18 Mar 2002, McCarthy, Doug FIREWALL wrote:

 > I wish I could be more specific.  I really do.  But when you see the

 > Mar 19 01:38:57 clio mimedefang[4542]: Could not create
 > /var/spool/MIMEDefang/mdefang-3C9696AF-0: File exists

whoahh... looks bad...

It's a long shot, but is it possible you have more than one multiplexor
process running?  That would cause Bad Things to happen.

Second: How reliable is the pthread implementation on your platform?
A flaky threads implementation could cause problems like this.

I've never seen anything like your messages before.  Anyone?

Regards,

David.
**********************************************

I checked my host and indeed there are 2 multiplexors, I killed one and the 
timeout error disappeared right away.  I stayed up all night checking the 
log, timeout error never showed up again.

But this morning I realized that nothing has been scanned!  I have trophie 
as well, this means even virus scan has not been running.

When I try to restart the program using the startup came with 
TheMIMEDEFANGHOWTO and what I had modified,  it forks 1 multiplexor and 
mimedefang, the timeout error occurs right away.  But when you eventually 
could send the message, it is scanned.

I killed the mimedefang process and left the multiploxor alone, timeout 
went away, but again, nothing has been scanned.

system info:
Solaris 8 on Sparc E450 with 2 CPUs, 2GB RAM
Sendmail8-12.9
Mimedefang2.31
Spamassassin2.44
Trophie 2.12
TrendMicro VirusWall, standard version


starting program:

1. start trophie
  /usr/local/bin/trophie -D

2. start mimedefang user the startup script in the example dir

# Minimum number of processes to keep.  The default of 0 is probably
# too low; we suggest 2 instead.
MX_MINIMUM=1

3. start sendmail:
     /usr/lib/sendmail -L sm-mta -bd -q45m
      /usr/lib/sendmail -L sm-msp-queue -Ac -q30m
      /usr/lib/sendmail -Ac -qp2m


What I am not clear is:

1.Should I start trophie separately or just through MD (enabled when 
installed)?

2. Do both mimedefang and multiplexor need to be running with MD2.31?

3. Could it be installation problem?

Thank you for any help you can give.  If I can not get this to work within 
a day or 2, we will have to roll back to our old setting:
sendmailswitch+spamd which overloads our system.


Xiaoyan

At 10:58 AM 4/9/2003 -0400, you wrote:
>On Tue, 8 Apr 2003, Xiaoyan Ma wrote:
>
> > Thanks Tony, I added these as well. Now majordomo is working.  But I still
> > get the "try again later" error.
>
>Could we see excerpts from your mail log?  It's hard to diagnose without
>that info.
>
>Regards,
>
>David.
>_______________________________________________
>MIMEDefang mailing list
>MIMEDefang at lists.roaringpenguin.com
>http://lists.roaringpenguin.com/mailman/listinfo/mimedefang