too long indexing process
"gape" <[email protected]> Thu, 7 Aug 2003 21:59:27 +0200
| Newsgroups | gmane.comp.web.perlfect-search |
|---|---|
| Message-ID | <002c01c35d1e$6bba4f40$5a0064c0@gapelt> |
i have a little problem i have perlfect 3.30 installed for a year now ... at first i was using it for indexind thru disk (document_root & stuff) then i changed the conf to index thru web (base_url & stuff), becouse of a forum ... i was indexing thru web as i remember it took perlfect about 30 mins to index the site afcourse the site has grown ... it got a php script geeklog where i run a 'news paper'. it has abot 1000 articles, some with comments. forum has 50000 posts when the index got thru (1 year ago) i gave the directive to TheMasterOfTheWebMachine, to put perlfect in cron. he didn't ... so ... now i wanted to finally, really put together a search engine, that will tell a searcher the ansver to the question. so i gave another directive to TheMasterOfTheWebMachine to index my site thru ssh and to look for processor times and so on ... so he did processor 100% all the time, the sites on machine were working, but ... 100% ??? sites were slower afcourse ... but that is all ... the machine is not experimental, it runs a lot of commercial sites. after a few hours TheMasterOfTheWebMachine called me, to tell me that he will stop the script, couse it ate 250+ of ram and all of the procesor ... i have $HTTP_MAX_PAGES = 10000; the first q is if i have $HTTP_START_URL enabled (uncomented), should i coment out $DOCUMENT_ROOT - now it wasn't. the second ... is it normal for indexer to run so long? i mean ... site has well over 10.000 pages (html, php, pl ...) ... the script should stop when it reached that limit ??? or what ... tnx ... With Love and Light gape www.gape.org