"Eyal Lanxner" <eyall@census.cbs.gov.il> wrote in message
news:0c7e01c36bde$cd4ca590$a501280a@phx.gbl...
> Hello,
>
> Our project involves a lot of data transformation,
> processing and matching operations on multi-versioned data
> sources (physical tables) - each containing about 10
> million records. The team for developing and executing
> these operations includes 15 people that are working on
> the data, each one on a version of his own. In order not
> to "choke" a single database server, the configuration
> will include a number of local databases for developing,
> executing and testing the needed procedures, and a central
> database that will integrate all the final processed data.
>
> My questions are:
>
> 1) Is there a product (MS or other) that implements
> version control on database tables (and not on files like
> source safe)? Can this product synchronize the control
> status of the data sources across all the databases? For
> example, if there is a data source which is localized in
> the central database, and there is a need to perform a
> major operation on it, it will be checked out into a local
> database, processed, and then checked in back in the
> central database. In this case, the control status of the
> data source should be known across the whole system, so no
> one updates it while it's locked.
>
> 2) What is the best database topology to use here? From
> one hand, the databases are physically located in
> different boxes; from the other, they all combine one
> logical database. Is clustering relevant here?
>
> Thanks in advance,
> Eyal Lanxner.