> >
> > I have a web app to allow admins to search this table. Any searches by
> > MsgDate, MsgTime, MsgPriority or MsgHostname or any combination of them
> > runs
> > pretty fast. When I try to do a search against MsgText, it takes quite
> > a
> > while. I originally tried INSTR(), however Explain showed that no
> > INDEX
> > was
> > being used. Using Match() Against() showed the Kiwi_MsgText FULLTEXT
> > index
> > getting used.
> >
> > What can I do to speed this up? Should I do a regular INDEX? I used
> > the
> > configuration setting from the example huge.cnf file.
One thing that ocurs to me is that, if you are using it for syslog type
data, the data being logged will have a relatively samll "vocabulary". No
tesxt-based indexing system will handle this very well. Like looking up a
very common name in a telephone directory, you will get a lot of hits which
will then have to be searched linearly for the right one. And linear search
of meven a subset of a buig database is bad news, because the records will
be physically scattered and not amenable to caching.
*If* it is true that your data is highly repetitive, could you arrange to
factor out the repetitive bits? For example, if there ate only 20 record
types can you parse out the type and convert it to an integer?
Alec
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
