Back to the month index |
Back to the list index
|
Vivek Khera (khera@kci.kciLink.com)
Wed, 08 Jan 1997 15:13:25 -0500
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
- Next message: Wayne Dahl: "[mSQL] bookmarks example"
- Previous message: Tom Henry: "Re: [mSQL] Advice on msql WWW-frontend"
- Maybe in reply to: Thomas Hintze: "[mSQL] Advice on msql WWW-frontend"
Message-Id: <199701082013.PAA07232@kci.kciLink.com> Subject: Re: [mSQL] Re: 2.0 - can it cache results? Date: Wed, 08 Jan 1997 15:13:25 -0500 From: Vivek Khera <khera@kci.kciLink.com>+---------
| I know many of the big search engines cache result tables, so that's
| another source of information. Might as well find out what's been done,
| avoid reinventing the wheel and so forth.
+---------
I use the Nextrieve (http://www.nexial.nl/) full-text engine, and when running
in server mode, it caches the recent queries. It can then use the cached
query to perform other similar queries. The cache expires after a certian
time or when the space is needed for other queries.
For example, if I query the data looking for items in cateogory "R" and having
a date > 19960509, then I query again for cateogory R with date > 19960910 and
keyword "population explosion", it can use the prior results. Typically a
search across the data takes something like 5 seconds if it is freshly loadad,
and subsequent queries take under a second.
However, it cannot handle arbitrary SQL-type queries. It only supports
attribute values which you enumerate ahead of time, which works great for
classifying documents, but not for raw data in general.
What I see as being necessary is either a middle-ware persistent data server
to which you submit queries rather than directly to the mSQL server. This new
server would cache the results and decide if your current query is similar to
one it already has in its cache and then just return the necessary values.
Otherwise, it will submit the query to the real database server and then cache
those results.
At some point building in such a cache to the main mSQL server would be handy,
too.
v.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Vivek Khera, Ph.D. Khera Communications, Inc.
Internet: khera@kciLink.com Rockville, MD +1-301-258-8292
PGP/MIME spoken here http://www.kciLink.com/home/khera/
--------------------------------------------------------------------------
To remove yourself from the Mini SQL mailing list send a message containing
"unsubscribe" to "unsubscribe" to msql-list-request@bunyip.com. Send a message containing
"info msql-list" to majordomo@bunyip.com for info on monthly archives of
the list. For more help, mail owner-msql-list@bunyip.com NOT the msql-list!
- Next message: Wayne Dahl: "[mSQL] bookmarks example"
- Previous message: Tom Henry: "Re: [mSQL] Advice on msql WWW-frontend"
- Maybe in reply to: Thomas Hintze: "[mSQL] Advice on msql WWW-frontend"