Very Computer

I would appriciate it if someone could advise/hint me on this issue.

Thanks

Jorg.

backgr. info:
Why you may wonder? I'm working on a program in Visual C++ which has 5
processes that store information in the database
(concurrent access). The main issue is to get a quick database for these
processes, without duplicates of the database in memory.
It's information in the RDBMS is mostly process information therefore the
database doesn't contain huge amounts of data.
I'm still considering which database to use for this job. It has to be a
fast one, perhaps SOLID Server, Raima Velocis, Pervasive.SQL  or
Sybase SQL Anyware.  You make my day with every comment, any suggestion?

Hi Jorg,

For what you describe, I think any of the databases would do a fine job.
Since the volume of data is small, none of the databases should bog down
due to volume of data.

If you need real-time or near real-time performance, we (Raima) might
have an edge because of the option of using the pointer-based network
model as opposed to relational model.  Inserts and retrievals will be
faster, predictable, and constant regardless of the amount of data
inserted.  Relational databases, because of the necessity to maintain
indexes, will be slower on inserts, unpredictable, and not constant
(slower as the database size grows).

The Velocis server uses its own internal lightweight thread scheduler.
It doesn't depend on the OS for multi-threading.  On single processor
machines with multiple clients (or threads within a single client), this
is a more efficient architecture than OS threads (our specialized
scheduler can task switch faster than the OS scheduler).

Jorg, email me directly if you want to discuss further.

Best regards,
Steve

Polyhedra is an Object-relational MMDB designed for high performance C/S
process control systems and has the features you need including interfaces
to PLCs and RTUs as well as high speed Historian for persistent data
logging/storage.  Check www.polyhedra.com

Dave Morse

Quote:>When you have multiple (almost realtime) processes (without forms) which
>have can access/manipulate a
>database locally on an NT server, which database client tech. is most
suited
>for this.
>I am trying to compare the different database client technologies to
>determine which client tech. is most suited for the job.
>I'm not sure though which is the best choice; ADO, RDO, DAO or DAO.

>I would appriciate it if someone could advise/hint me on this issue.

>Thanks

>Jorg.

>backgr. info:
>Why you may wonder? I'm working on a program in Visual C++ which has 5
>processes that store information in the database
>(concurrent access). The main issue is to get a quick database for these
>processes, without duplicates of the database in memory.
>It's information in the RDBMS is mostly process information therefore the
>database doesn't contain huge amounts of data.
>I'm still considering which database to use for this job. It has to be a
>fast one, perhaps SOLID Server, Raima Velocis, Pervasive.SQL  or
>Sybase SQL Anyware.  You make my day with every comment, any suggestion?

Perhaps RDO (Remote Data Objects) is out of your scope because your database
is situated locally.

I think ODBC could be a good alternative. Especially MFC ODBC because you
are programming in Visual C++, although I don't have experience in
concurrent access to databases via ODBC, should be possible though, I guess.

Try to stay out of the hands of database vendors, who seemed to have
responded to your message, until you've made a good pre-eval of the
databases currently at hand.

So far my 2 cent advice.

Carl Dunmore.

Quote:> When you have multiple (almost realtime) processes (without forms) which
> have can access/manipulate a
> database locally on an NT server, which database client tech. is most
suited
> for this.
> I am trying to compare the different database client technologies to
> determine which client tech. is most suited for the job.
> I'm not sure though which is the best choice; ADO, RDO, DAO or DAO.

> I would appriciate it if someone could advise/hint me on this issue.

> Thanks

> Jorg.

> backgr. info:
> Why you may wonder? I'm working on a program in Visual C++ which has 5
> processes that store information in the database
> (concurrent access). The main issue is to get a quick database for these
> processes, without duplicates of the database in memory.
> It's information in the RDBMS is mostly process information therefore the
> database doesn't contain huge amounts of data.
> I'm still considering which database to use for this job. It has to be a
> fast one, perhaps SOLID Server, Raima Velocis, Pervasive.SQL  or
> Sybase SQL Anyware.  You make my day with every comment, any suggestion?