[Mimedefang] No free slaves and high CPU usage

Steffen Kaiser skmimedefang at smail.inf.fh-bonn-rhein-sieg.de
Thu Jul 3 09:28:34 EDT 2008


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thu, 3 Jul 2008, Jan-Pieter Cornet wrote:

>> "mimedefang[6136]: [ID 669130 mail.warning] mfconnect: No free slaves:
>> Need 1, found 0"

Well, this is normal, because if all your slaves are already working, you 
cannot allocate another one.

>> Under /var/spool/MIMEDefang I can see around 26710 directories that
>> starts with  "mdefang-m<random_numbers_and_characters>" - most of
>> which are empty.

Hmm, I had something like that myself, it turned out that I had a 
debugging flag still on the production server, "-k" to mimedefang, see man 
page, "Causes mimedefang not to delete working directories if a filter 
fails." I don't know if this "no free slave" is such a condition.

> That's bad. I suggest you shut down sendmail, then delete all those
> mdefang-* directories, and restart sendmail again (if you want to
> be thorough, also shut down mimedefang, and recreate the
> /var/spool/MIMEDefang directory itself, because it likely grew
> quite big).

Yep.

> Also - if your /var/spool/MIMEDefang is not on a RAM disk, I want to
> urge you to store it on RAM disk. There is nothing stored there that

Yep again.

Bye,

- -- 
Steffen Kaiser
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFIbNQF5ThHZhj8SBwRAqAVAJ9nohpsi86s5CPLcgPrmUVddLqQewCaA89V
IjPze/hTR807vUgV04RuK7o=
=kakW
-----END PGP SIGNATURE-----



More information about the MIMEDefang mailing list