You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@spamassassin.apache.org by bu...@bugzilla.spamassassin.org on 2004/07/28 18:19:00 UTC

[Bug 3641] New: make spamd debugging easier

http://bugzilla.spamassassin.org/show_bug.cgi?id=3641

           Summary: make spamd debugging easier
           Product: Spamassassin
           Version: SVN Trunk (Latest Devel Version)
          Platform: Other
        OS/Version: other
            Status: NEW
          Severity: enhancement
          Priority: P1
         Component: spamc/spamd
        AssignedTo: spamassassin-dev@incubator.apache.org
        ReportedBy: dallase@nmgi.com


i've been trying to get to the bottom of running spamd children that never seem 
to die.

31833 root      20   0 29576  28M  2252 R    67.3  3.8 296:28 spamd
21845 root      19   0 27164  26M   120 R    78.9  3.5 320:48 spamd

while this is very hard to reproduce as it happens about 2 or 3 times a day out 
of 250k emails, its even harder to debug when there is that much mail flow.

i propose adding the process id to ever dbg() call so its easier to grep for a 
the PID of a spamd child in the logs to show all debug info for that child... 
because with multiple childs, the lines get tangled when they are all 
processing at the same time. 

# grep "\[2368\]" /var/log/spamd/spamdlog
debug: [2368] running uri tests; score so far=-1.384
debug: [2368] running raw-body-text per-line regexp tests; score so far=-1.384
debug: [2368] running full-text regexp tests; score so far=-1.384
debug: [2368] auto-learn? ham=0.5, spam=7, body-points=0, head-points=-2.79
debug: [2368] auto-learn: currently using scoreset 1.  no need to recompute.
debug: [2368] auto-learn? no: learner indicated spam (1.6965 > 1)
debug: [2368] is spam? score=0.312 required=4
debug: [2368] tests=ALL_TRUSTED,AWL,MISSING_DATE,MISSING_SUBJECT
debug: [2368] subtests=__UNUSABLE_MSGID

i also propose setting the razor debug level to 0 because that accounts for 
nearly 20% of the debugging info.   also, make the tokenizing and bayes token 
debug commands honor $DEBUG->{bayes}=0;  - or maybe create a $DEBUG->
{bayes_toks}=0; and make them honor that... because they tend to produce alot 
of debug info.

the following patch that i will submit does these things.. i hope you will 
consider :)



------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.