dupe BE just for Reports?

[GBA]

Have an abstract question to put out there for friendly commentary. If one
has two core classes of users where one is putting in data and the other is
looking only at reports....

and those reports are big trend reports that suck in alot of tables....

Is there any positive rationale to the idea of duplicating the BE (named
maybe BE-Reports ) and having the FE of the report readers go to the
duplicate BE-Reports in order to offload the main BE ??

Response time for the main users so far isn't bad...a couple seconds to open
some forms or change records. Kind of mulling this in the theoretical
sense. The report readers don't need the current day's info...so I could
dupe it daily for them and segregate the two classes of users...(assuming I
could figure out some system code to do that task automated...)

Would welcome friendly comments on the concept...have never seen discussion
on this idea but maybe because there isn't usually such a firm class
distinction in the role of the users. But in my case it is pretty clear &
firm where the report readers really do not data entry at all and could have
their own Fe/Be..

[Jeff Boyce]

There are so many factors that can impact this decision that you're probably
better off testing YOUR situation to see ...

For example...
* how many users (inputting, reporting)?
* which version of Access?
* speed of network?
* back-end data stored in what (?Access?, ?SQL-Server?, ...)?
* indexes on fields in tables?
* are you using queries to 'feed' the forms?
* are you using queries to 'feed' the reports?

... and that's just a start!

Good luck!

Regards

Jeff Boyce
Microsoft Office/Access MVP

"GBA" <G...@discussions.microsoft.com> wrote in message
news:DAF9568E-0349-42D6...@microsoft.com...

[david]

Not normally required, but that doesn't mean never.

SQL Server is optimised to keep a lot of things in
memory, to give reasonable service to a lot of people
doing a lot of different things, using huge databases.
When one user uses all of the memory and processor
power to create a huge recordset, everyone else stops.

Access/Jet uses a completely different model. The database
can only be 2GB in size. The indexes are copied down to
the user and placed in user memory. The joins are done in
user memory. The join indexing is done by the user processor.
The whole distributed processing system is markedly more
tolerant.

If one user runs a massive report using all the data, in a
shared jet/mdb database, depending on what they are doing,
other users may not even notice.

(david)

"GBA" <G...@discussions.microsoft.com> wrote in message
news:DAF9568E-0349-42D6...@microsoft.com...