Hi A few thoughts. I am not an expert on this kind of use, so there may be others with better suggestions :)
If every search is different then your search efficiency is critical. If you fulltext index the field then it should help. If not then what about indexing the text document words. You scan the table and index every word except the linking ones (and a so etc) eg WordIndex Table DocID | Word | WordCount| -------------------------- 123 | zip | 2 | 123 | zap | 3 | 144 | zip | 4 | 973 | zip | 8 | 973 | zap | 4 | This allows better searching too eg if the search is for 'zip' then Doc 973 is more relevant than Doc 123 This might be a better long term and more scalable answer more hardware power. Just my 2p worth Peter ----------------------------------------------- Excellence in internet and open source software ----------------------------------------------- Sunmaia www.sunmaia.net [EMAIL PROTECTED] tel. 0121-242-1473 ----------------------------------------------- > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] > Sent: 27 June 2002 18:35 > To: [EMAIL PROTECTED] > Subject: Hardware Planning > > > I am currently using mysql 4.01 with caching for web page serving. I > pull several records out with a like statement on a text field. The db > consists of about 200,000 records right now and it is using 100% > of the cpu > when the search is ran. > Since this is a critical function, it is ran on almost every web > request. Without being cached, the search can take up to 2000-3000 > milliseconds. With cache it is around 40-50 ms, but there are so many > different queries, cache is almost useless. It is killing the > machine though > as nothing else can be ran on it. > > The machine specs: > Windows 2000 > Pentium III 733 MHZ > 756 SDRAM > SCSI Drives. > MYSQL++ API > > I am looking into having the records approach 1 million records soon and > would really like to minimize cpu and search time. Any > suggestions on a new > machine or does this search time and cpu usage not sound right? Thanks in > advance. > > Jeremy Beha > > > --------------------------------------------------------------------- > Before posting, please check: > http://www.mysql.com/manual.php (the manual) > http://lists.mysql.com/ (the list archive) > > To request this thread, e-mail <[EMAIL PROTECTED]> > To unsubscribe, e-mail > <[EMAIL PROTECTED]> > Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php > --------------------------------------------------------------------- Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail <[EMAIL PROTECTED]> To unsubscribe, e-mail <[EMAIL PROTECTED]> Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
