Menu
▾
▴
#678 0.21.1 Cross Platform - Bayes.pm and DBI::SQLite errors.
Please forgive my vagueness with this one, but I ran
into over 15 errors that occurred while running POPFile
0.21.1 within the first 30 minutes after upgrade from
0.18.3.
Almost all of the errors point towards
"/Classifier/Bayes.pm" and the DBI::SQLite.
I have included them in the attched console log file
since I was unable to paste it all into the window.
As for the configuration information, it is the same as
I have most recently referenced in my post under
POPFile Bug Report 917994.
Also, on an upside, POPFile has been running (filtering
email) for over a day now without any crashes. So,
thankfully it is very stable in that regard. It's just
the interaction with the web interface that crashes
POPFile at times.
POPFile 0.21.1 Cross Platform Console Log
Logged In: YES
user_id=171106
>DBD::SQLite::st execute failed: library routine called out of
sequence at /Classifier/Bayes.pm line 2871, <$reader> line 5.
Those are some really interesting errors, what OS are you
running? OSX?
What version of dbd::SQLite do you have installed?
Scott
Logged In: YES
user_id=951169
I currently have DBD-SQLite-0.31 installed on my Red Hat Linux 7.3 box.
However, I admin the software and use my email client on Mac OS
10.3.3, but I don't believe that would be causing it to generate those
previous errors.
I'm no longer running the software in a terminal window. I have it
running as a daemon. I'm not sure if it was some weird, run it a few
times to clear out the problems kind of a thing or what. Because the
other dozen times I did start up and use POPFile via a terminal window, I
didn't see any other error messages. I'm going to fire it up in a terminal
window now though. That way I can see if anything is being generated
that doesn't make it to the POPFile logfile. I'll post a reply with my
findings. Thanks, Steveo
Logged In: YES
user_id=558706
I'm experiencing the same behavior using POPFile 0.21.1a for
Windows, on Windows XP. I had an install of 0.20.1 (I
believe) before installing this over it. The error behavior
started when I attempted to reclassify a message about half
an hour ago, and has happened consistently since.
Here's what happens when I try starting up POPFile manually,
after it tries to start up when the computer boots and then
exits after a minute or two. It occurs when I run POPFile,
not just when trying to load the web interface.
POPFile Engine loading
Loading...
{core: windows}
{core: mq logger config}
{classifier: bayes wordmangle}
{interface: html}
{proxy: pop3}
POPFile Engine v0.21.1 starting
Initializing...
{core: mq windows config logger}
{interface: html}
{classifier: wordmangle bayes}
{proxy: pop3}
Starting...
{core: mq windows
A copy of POPFile appears to be running.
Attempting to signal the previous copy.
Waiting 10 seconds for a reply.
The other POPFile (4060) failed to signal back, starting new
copy (352)
config logger}
{interface: html}
{classifier: wordmangleUse of uninitialized value
in subroutine entry at d:\popfile/Classifier/Bayes.pm line 733.
DBD::SQLite::st execute failed: at
d:\popfile/Classifier/Bayes.pm line 733.
Use of uninitialized value in subroutine entry at
d:\popfile/Classifier/Bayes.pm line 744.
DBD::SQLite::st execute failed: at
d:\popfile/Classifier/Bayes.pm line 744.
Use of uninitialized value in subroutine entry at
d:\popfile/Classifier/Bayes.pm line 2683.
DBD::SQLite::st execute failed: at
d:\popfile/Classifier/Bayes.pm line 2683.
Can't use an undefined value as an ARRAY reference at
d:\popfile/Classifier/Bayes.pm line 2684.
Logged In: YES
user_id=951169
Scott,
After running POPFile 0.21.1 Cross Platform for over a day, I've yet to
see any more of the Bayes error messages, or any other error message
at all. I think those error messages was from some initial launching of
POPFile after it upgraded my old version of POPFile. All I know is that the
plumbing is clean, and that if there is an issue it appears to only be a
temporary one. I've since restarted the Red Hat Linux box several times,
and everything is running great. All except for POPFile's "Broken pipe"
issues, but that's a totally different issue that I've already reported.
Thanks, Steveo.
Logged In: YES
user_id=171106
Steveo,
Glad to hear it cleared up. Bothers me that it happened at
all, but in absence of more clues, it will be hard, if not
impossible, to nail it down.
As for the broken pipes, any chance that they are happening
when you click the stop button on the browser? I've seen
reports elsewhere that some browsers send a broken pipe
signal on stops.
Scott
Logged In: YES
user_id=897713
"Me too." I'm seeing similar problems running 0.21.1 with
DBI::MySQL on OSX. After running for a period of time
sorting mail, a reclassify crashes popfile. MySQL continues
to run. Restarting popfile, I see the reclassification
occurred, but the ID's start at a very large number. The
weird numbering clears up after receiving more mail.
I have also seen the error messages trying to access the
database, but it's not right in front of me to give a
precise report.
Logged In: YES
user_id=897713
To expand on my last comment, when popfile crashes, I get a
string of error messages. The MySQL server (4.0.18) seems
to be fine, and I just need to restart popfile. Hope this
helps:
DBD::mysql::st execute failed: MySQL server has gone away at
/Classifier/Bayes.pm line 2821, <$reader> line 3.
DBD::mysql::st fetchrow_arrayref failed: fetch() without
execute() at /Classifier/Bayes.pm line 2822, <$reader> line 3.
DBD::mysql::st execute failed: MySQL server has gone away at
/Classifier/Bayes.pm line 2828, <$reader> line 3.
DBD::mysql::st fetchrow_arrayref failed: fetch() without
execute() at /Classifier/Bayes.pm line 2830, <$reader> line 3.
Use of uninitialized value in addition (+) at
/Classifier/Bayes.pm line 290, <$reader> line 3.
DBD::mysql::st execute failed: MySQL server has gone away at
/Classifier/Bayes.pm line 2871, <$reader> line 3.
DBD::mysql::st execute failed: MySQL server has gone away at
/Classifier/Bayes.pm line 2821, <$reader> line 5.
DBD::mysql::st fetchrow_arrayref failed: fetch() without
execute() at /Classifier/Bayes.pm line 2822, <$reader> line 5.
DBD::mysql::st execute failed: MySQL server has gone away at
/Classifier/Bayes.pm line 2828, <$reader> line 5.
DBD::mysql::st fetchrow_arrayref failed: fetch() without
execute() at /Classifier/Bayes.pm line 2830, <$reader> line 5.
Use of uninitialized value in addition (+) at
/Classifier/Bayes.pm line 290, <$reader> line 5.
DBD::mysql::st execute failed: MySQL server has gone away at
/Classifier/Bayes.pm line 2871, <$reader> line 5.
Logged In: YES
user_id=558706
The errors I described earlier seem to have resulted from a
bad cluster in my popfile.db database - I get I/O errors
when trying to copy it. Moving it away and regenerating it
fixes the problem. So it doesn't seem to be related after all.