| | |
Need authoritative source on safety of resetting ft_min_word_len
Please support our MySQL advertiser: PostgreSQL or MySQL? Compare and contrast the two most popular open source databases
![]() |
•
•
Join Date: Jul 2006
Posts: 1
Reputation:
Solved Threads: 0
I need to set ft_min_word_length=3, so that my users can perform boolean fulltext searches for words of three characters instead of four. However, I share the MySQL server with other customers of my ISP.
My ISP's administrator is willing to change this system variable, provided I can show him an authoritative source that says doing so won't break anything if other customers don't rebuild their FULLTEXT indexes.
The description on http://dev.mysql.com/doc/refman/5.1/...ne-tuning.html contains no warnings; the text only implies that setting ft_min_word_len=3 will have no effect until you rebuild your FULLTEXT indexes (which should continue to work as they did before).
The following quote from my admin's latest email explains his problem:
Anybody know of an authoritative source that explains the consequences?
My ISP's administrator is willing to change this system variable, provided I can show him an authoritative source that says doing so won't break anything if other customers don't rebuild their FULLTEXT indexes.
The description on http://dev.mysql.com/doc/refman/5.1/...ne-tuning.html contains no warnings; the text only implies that setting ft_min_word_len=3 will have no effect until you rebuild your FULLTEXT indexes (which should continue to work as they did before).
The following quote from my admin's latest email explains his problem:
•
•
•
•
Can you find any authoritative source that actually says that the only consequence is that text searches won't find any 3-character words because the fulltext index won't contain them until it's rebuilt? The MYSQL manual doesn't clearly say this, and it does say that "you must rebuild your fulltext index."
There are quite a few other MySQL databases on the server, many of them automatically created when installatron is used, and I have no idea if any of them use fulltext searches. I can't afford the risk of any of them breaking. So all I need is some authoritative source that says that existing databases will not be affected.
![]() |
Similar Threads
- Windows 2K source code? (Windows NT / 2000 / XP)
- Can't view source codes (Web Browsers)
- Beware of Bad Capacitors (source of much bizarre behavior) (Motherboards, CPUs and RAM)
- View Source in IE6 (Web Browsers)
- PCModKingdom Source Code (IT Professionals' Lounge)
Other Threads in the MySQL Forum
- Previous Thread: Mysql Error
- Next Thread: Help With Messages in Mysql Stop Overwriting
| Thread Tools | Search this Thread |
agplv3 alfresco amazon api aws bizspark breathalyzer changingprices cmg communityjournalism contentmanagement contractors copyright count court crm database design developer development distinct drupal dui ec2 email enterprise eudora facebook form foss gartner gnu government gpl greenit groklaw groupware hiring hyperic images innerjoins insert ip joebrockmeier join journalism keyword keywords kickfire laptop law legal license licensing linux maintenance managing mariadb matchingcolumns metron micromanage microsoft microsoftexchange mindtouch montywidenius mozilla multiple mysql mysqlcolumnupdating mysqldatetimeordermax() mysqlindex mysqlinternalqueries mysqlquery mysqlsearch news open-xchange opendatabasealliance opengovernment opensource oracle penelope php priceupdating query referencedesign reorderingcolumns resultset saas select sharepoint simpledb sourcecode spotify sql sugarcrm syntax techsupport thunderbird transparency virtualization





