Hi there again,
Scott Kitterman:
> On Monday 07 July 2008 13:41, Zbigniew Szalbot wrote:
>> Hi again,
>>
>> Scott Kitterman:
>>> On Monday 07 July 2008 13:33, Zbigniew Szalbot wrote:
>>>> Hello,
>>>>
>>>> Scott Kitterman:
>>>>> On Monday 07 July 2008 12:40, Murray S. Kucherawy wrote:
>>>>>> On Mon, 23 Jun 2008, Zbigniew Szalbot wrote:
>>>>>>> Jun 21 20:02:43 relay postfix/cleanup[39722]: warning: milter
>>>>>>> inet:127.0.0.1:4445: can't read SMFIC_BODYEOB reply packet
>>>>>>> header: Unknown error: 0 Jun 21 20:02:43 relay
>>>>>>> dkim-filter[783]: terminated with signal 11, restarting Jun
>>>>>>> 21 20:02:43 relay dkim-filter[39913]: Sendmail DKIM Filter
>>>>>>> v2.6.0 starting (args: -l -p inet:[EMAIL PROTECTED] -u mailnull
>>>>>>> -P /var/run/milterdkim/pid -x
>>>>>>> /usr/local/etc/mail/dkim-filter.conf)
>>>>>>>
>>>>>>> Unfortunately, I no longer have this message. Is there a
>>>>>>> configuration setting that would let me know about service
>>>>>>> restart (by email for example)? It would make it easier to
>>>>>>> check for offending message. Thanks!
>>>>>> This is a known problem. I've looked into it with the help of
>>>>>> some other list participants, but so far I don't have a
>>>>>> solution. Oddly, it only appears on systems where Postfix is
>>>>>> the MTA in use.
>>>>>>
>>>>>> The forensic data I've been shown suggests that something is
>>>>>> closing the socket that connects the filter to the MTA
>>>>>> improperly, but it's very hard to determine where or how that's
>>>>>> happening.
>>>>> I have a vague recollection of this coming on up postfix-users
>>>>> and being related to assumptions about buffer sizes that were
>>>>> unfortunate and it being fixed, but have been unable to find it
>>>>> in the archive. IIRC the OP is on Debian. There are newer
>>>>> Postfix packages on backports.org. You ought to see if you can
>>>>> replicate the problem with a newer Postfix.
>>>> Not sure where from you gather I am on Debian :), because I am a
>>>> happy FreeBSD user, but this is not important here.
>>>>
>>>> I am using postfix-2.5.1_2,1, which I think is pretty much the
>>>> latest version available.
>>> Sorry about that. I must have confused the threads. I haven't seen
>>> anything like that myself since 2.3.
>> No need to be sorry! :) Do you think it is worthwhile to bring this up
>> on a postfix list to see what wise people there may have to say?
>>
>> For me this is normal that DKIM gets restarted every 2-3 days:
>>
>> $ grep restarting /var/log/maillog
>> Jul 7 09:00:55 relay dkim-filter[787]: terminated with signal 11,
>> restarting
>>
>> $ gzcat /var/log/maillog.2.bz2 |grep "restarting"
>> Jul 4 03:23:16 relay dkim-filter[787]: terminated with signal 11,
>> restarting
>>
> I think so. Be sure to supply all the information asked for in the
> postfix-users welcome message (also in the debug README).
I do not think I have much to offer from the postfix list...
"Signal 11 is a bug in the program that dies. It is usually the
result from an attempt to access an invalid memory address.
As for how to debug dkim-milter, this is not the most appropriate
mailing list. All I can do is tell you to record all postfix-to-milter
traffic with tcpdump including the full packet content, and if you
can capture a session where dkim-milter crashes, then I can tell
you what the packets meant.
Wietse"
Capturing a session might be quite a challenge as it crashes randomly at
any time, usually once every 3-4 days...
Thanks!
--
Zbigniew Szalbot
www.LCWords.com
-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
dkim-milter-discuss mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dkim-milter-discuss
