clustered. Built a doc and image library. Had to put these files on
a special share so all members of the cluster have access to them. A
Celera SAN thingy actually, works fine and dandy, but one more system
to depend on for this.
What about speed? If say I get 50 requests for a 1MB document, what
would be speedier? fetching from the DB or fetching from the file
system?
Oracle Text? Part of 9i?
Doug
----- Original Message -----
From: James Holmes <[EMAIL PROTECTED]>
Date: Sat, 25 Sep 2004 00:46:46 -0400
Subject: Re: BLOB
To: CF-Talk <[EMAIL PROTECTED]>
All of these are reasons we use the DB for all dynamically managed
content, including binary files. Oracle is absolutely fantastic at
this these days. For example, Oracle Text is far superior to the
Verity implementation in CF, since the full text query can be included
in the SQL statement used to retrieve doc info (including version info
and many other things that are more difficult to manage with content
written to a filesystem). It also means that if multiple CF servers
are used or a DR server is brought online, we don't have to worry
about where Verity files were written (the index is all in the DB,
with the content, where it belongs). A few years ago BLOBS were bad
but things have come a long way.
>multiuser, continuous data, data integrity, transactions, business rule
>implementation, centralized storage and management, not having to manage
>millions of files, etc. pretty much every reason why you'd use a db in the
>first place. the whole GIS industry is running its enterprise databases on
>modern BLOB (spatial) technology. ibm, oracle, ms, esri, mapinfo, etc. have
>all spent a bundle on this technology. even freebies like postgres (postGIS
>is actually pretty slick) & mySQL are on the
bandwagon.________________________________
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings] [Donations and Support]
