Splitting Database over 2 countries

  • Thread starter Thread starter Guest
  • Start date Start date
G

Guest

I have a database that need to run on two sites, one in Germany and the
second in Britain. Both sites are connected on the in-firm network. I have
tried splitting into Frontend and Backend with the Backend stored on the
local network at one of the sites but the time delay in starting the Frontend
from the site without a Backend is in the minutes range which of course means
that nobody will use it.
I have also thought of using replication but there are a number of features
in the Database which necessitate the deletion and creation of files during
usage and as these are design changes they are not permitted on the
replicated database.
I've run out of options now but I'm hoping that somebody else may have some
suggestion.
Thanks.
 
When you write "deleting and creating files" do you really mean tables? If
you are creating and deleting files then they are external to the database
and therefore not an issue where replication is concerned. If you mean
Tables then you can use delete queries to clear out the table(s) before
using append queries or other means to fill them.

Maybe you could explain to us, so we can better help you, how it is that
each of two sites might delete and/or recreate information that the other
site might also delete and/or create> I realize I'm assuming a lot but you
haven't told us much.

HTH
 
I wrote an application for a company a couple of years ago that they
insisted would only be used in their main operation. Three months after I
deployed it they decided that they wanted their remote sites to use it too.
I tried the WAN thing as well, but it did not work any better than your
experience.

I looked into replication, but had no experience in that arena and they
wanted it "NOW", so I discarded that idea. Although I used a similiar
concept. I gave the remote sites a slightly different version of the
application than the main office. Each office maintained their own data,
and refreshed the main database via ADO connections (rather than linked
tables) several times a day (AM and PM). The startup code at the remote
sites checked the tables that were maintained by the home office (each table
contained a LastUpdated field so the remote sites could update those tables
on their system that had been changed at the home office during the previous
day). None of the data in these tables was time sensitive, so a one day lag
was not critical. I also had a Timer event setup that checked to see if the
application was idle and maintained a record of whether any local data had
change or been entered during the previous timer interval (this was set to
30 minutes of something like that). At the end of that interval, the system
would use ADO connections to update the tables in the main offices database.

This worked well and since I was only transmitting changed and new data each
direction.
 
Larry,
thanks for your reply.
Sorry, yes I did mean tables when I said files. Or rather in this case a
Query. I have a search function where the user can choose any combination of
five fields which uses a Query definition.
e.g.
Set qdef = DB.CreateQueryDef("Search_Query_Insert", "select * into
Inserts_Extras from Saved_Query_Insert_Extras where " & search1 & " and " &
search2 & " and " & search3 & " and " & search4 & " and " & search5)
where the fields search1 to search5 represent the choices made by the user.
Each time I set the Query definition I have to delete the older version of
the query which is where I'm having difficulties as this is classed as a
design change.
I hope this has cleared up any ambiguities.
I suppose one way around would be to do away with the query call and write
it out using RunSql in the Visual Basic editor?
Thanks again
Alan
 
I don't know anything about splitting databases, but you could avoid creating
and deleting queries by leaving the query as a saved query, and changing the
SQL property of the saved query from code as needed.

John
 
Back
Top