Re: data storage for finite element applications
Date: Wed, 14 Mar 2007 09:38:14 GMT
Message-ID: <aoPJh.10983$8U4.3613_at_news-server.bigpond.net.au>
"jma" <junkmailavoid_at_yahoo.com> wrote in message
news:1173803305.631584.183110_at_j27g2000cwj.googlegroups.com...
> Hello all,
>
> I am facing the following task: I want to benchmark different
> solutions for storing large amounts of finite element data from
> different simulations into one file that preferably resides in a local
> directory. One option of course is to stuff everything into one binary
> file. The other option is to use a DBMS. There is also a hybrid
> option, namely to store binary chunks into a DBMS, however, I guess in
> this case I do not take advantage of any relational data structures. I
> find it hard to figure out why most vendors use inhouse "database"
> systems instead of widely used commercial ones. Do I miss something?
> Any info, experience or redirection will be highly appreciated!!!
How many people are going to need access to the data? How many rows of element data will exist in the biggest table? How much update is going to happen per hour in rows? How many concurrent uses will update at this rate? If DBMS is one option, an RDBMS is another. (eg: SQL Server, Oracle, DB2, etc) Received on Wed Mar 14 2007 - 10:38:14 CET