Re: Feasibility of LARGE text database
Date: Thu, 3 Mar 1994 17:51:02 GMT
Message-ID: <rbs-030394094132_at_fractaldata.com>
In article <CM297s.8K1_at_news.cis.umn.edu>, rediske_at_lenti.med.umn.edu (Ryan Rediske (BioMed Lib)) wrote:
> Hi all,
> I'm a curious student employee of the U of MN. We run expensive
> database software to search medical journals (MEDLINE). As the subject
> states, it is very large (6-7 Gigabytes!). I know nothing of Oracle,
> other than it is a database system. Could we save our department tens of
> thousands of dollars by using our site license for the database file to
> make our own front-end and an Oracle based database? Is it possible
> considering the frequency of users (estimated 2,000+ login's a day, not to
> mention several queries per login)? We have a SunSparc10 model 51 with
> one processor, 128 Megs RAM and 9.4 Gigs of drives.
Not really.
You obviously could store the documents in the db, and use the db to distribute them, but with 6-7 GB you need content based retrieval. Oracle has a text search product, and from what I've seen it works well; but, it ain't cheap. Considering the cost of application development & deployment, I doubt it would pay off.
-- Robert Stucky Education is what remains (415) 948-5583 voice, fax, ARA after one has forgotten rbs_at_fractaldata.com everything he learned in school. A. EinsteinReceived on Thu Mar 03 1994 - 18:51:02 CET