#124 pages get left ouf of index; random

need info
closed
nobody
None
5
2002-06-13
2002-06-12
No

Here's a problem I've been having since I've began using
htdig on my site (www.gomagazine.nl, running solaris
2.8 and the most recent Apache).

Htdig installed and compiled OK, the cgi works
beautifully.But: not all pages are in the index. And to
make things worse, sometimes they are.

scenario A
1. run bin/rundig -a
2. search for 'kenny larkin'

no results

scenario B
3. run bin/rundig -a
4. search for 'kenny larkin'

some relevant results

scenario C
5. run bin/htdig -a
6. search for 'kenny larkin'

the most relevant result (interview with kenny larkin on
top).

I run the rundig script every night from cron and I can tell
whether it has indexed all my pages when I receive an
email telling me that htdig couldn't find acroread. Which
is excellent, because that means htdig has FOUND
the .pdf and can't do anything with it (which I don't mind).

For this brief period I am able to find virtually every
document on my site.

The night after that, I receive no email and things are left
out again. I honestly cannot put my finger on it. I've run
htdig -vvvv a dozen of times and it's all abacadrba to me.
htmerge is deleting stuff, but I gather that's OK in some
cases.

I can provide you with anything you ask: -vvvv output,
config file, more system info, anything. Help is
appreciated :)

Discussion

  • Gilles Detillieux

    Logged In: YES
    user_id=149687

    Well, without anything more concrete to pin this down to a
    bug,
    I'd recommend treating this like a likely configuration
    problem,
    and as such it should be discussed on the htdig-general
    mailing list, which is a better forum for this sort of
    interaction
    than the bug database is. Also, any such problem report is
    pretty close to meaningless without any indication of which
    version of ht://Dig you're running. If you're not running
    3.1.6,
    give that version a try first, and if the problem persists
    post a
    followup on htdig-general.

     
  • Gilles Detillieux

    • milestone: --> need info
    • status: open --> closed
     

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks