According to Gerd Brost:
> I tried searching the archive for my problem, but i didnt find a solution.
> so here is the problem:
> When we run htdig, it can happen that htdig crashes without warning or error
> message. it just consumes 90% cpu load and stays in that state forever.
> The search doesnt work anymore because the db files are corrupted i guess.
The db files that htdig generates are not useable by htsearch until
htdig successfully runs to completion, and htmerge is run to finish
up what htdig produced. If you need more verbose output from htdig,
run with one or several -v options. With 4 or more of these, e.g.
htdig -i -vvvv, you should get a ton of debugging output that will
help you narrow down what htdig is doing just before freezing up.
If you run htdig from the debugger, you can kill it when it freezes
and get a stack backtrace, that would be extremely helpful as well.
I'm going under the assumption that this is a platform-specific
problem, as it hasn't been reported before, that I can recall.
Please let us know, along with the debugging results above, what
platform you're running on: OS version, distribution if applicable,
htdig version, which htdig binary package if applicable, compiler
version, and anything else you might think is relevant.
> We got 2 error messages in our httpd-error-log:
> DB2 problem: /.../htdig/common/german/word2root.db: no such file or dir
> Is this caused when htdig didnt run completely and word2root wasnt
> recreated?
> this file exists and permissions are fine.
> and:
> DB2 problem: /.../htdig/db/db.words.db: page 0: reference count overflow
> any idea on that one??
The word2root.db and root2word.db are generated by "htfuzzy endings",
which is normally run by the "rundig" script. These are unrelated to what
htdig produces, but are needed by htsearch if you have the "endings"
algorithm defined in your search_algorithm attribute setting. If the
file exists, yet htsearch can't find it, it's usually because:
1) the path to it is wrong in your config file for htsearch,
2) the file is not readable by the user ID under which the web server runs,
or the directories leading up to it are not searchable (i.e. executable)
by this user ID, or
3) the web server runs htsearch in a virtual root environment, so the paths
seen by the CGI program don't match those you see from your login shell
(this is pretty rare).
The reference count overflow on db.words.db suggests some database
corruption, which isn't surprising given the problems you've had getting
htdig to run to completion.
--
Gilles R. Detillieux E-mail: <[EMAIL PROTECTED]>
Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/~grdetil
Dept. Physiology, U. of Manitoba Phone: (204)789-3766
Winnipeg, MB R3E 3J7 (Canada) Fax: (204)789-3930
_______________________________________________
htdig-general mailing list <[EMAIL PROTECTED]>
Information: http://lists.sourceforge.net/lists/listinfo/htdig-general
FAQ: http://htdig.sourceforge.net/FAQ.html
