Re: Many Schemas

Brian Tkatch wrote:
>
> In a case where one company is providing a service for other companies
> (hundreds, possibly thousands). The same service for each one. Would
> it be better to give each company their own schema or to have one
> schema and use a foreign key to separate the common lists. Obviously
> there are many points to consider, but I do not know where to start.
>
> By separating them, I would assume that aggregate reports are harder
> to maintain, and longer to run. I am just guessing that accessing
> hundreds of small indexed tables would be slower than accessing one
> large one (indexed as well).
>
> The individual queries on the system would probably be faster, for the
> lesser amount of rows. And, updates are less likely to ruin the system
> by a careless mistake, because schemas must be directly referenced.
>
> Any *pointers would be appreceated.
>
> Brian

Check out Oracle's Advanced Security and Virtual Private Database. you could implement your own access_set methodology, whereby any query is appended with a where clause to limit access - independent of program accessing the database.

Paul Received on Sat May 12 2001 - 18:24:30 CDT