🚀 go-pugleaf

Georg,

I think you might be looking at a similar set-up to what I use.
I have a central database and a number of sites connected. The main database
is where everyone replicates to and from. I've added a user at every site
called REPL_XXX (XXX is a site code to make it simple for me to use)
Each site then has a replication condition to replicate where the user name
is not the repl user that sent the info. This way it replicates out from the
central database to all the others any changes made anywhere but not back to
its self.
This works well. Anytime I need to add a new database I just Sync a copy to
the new user, and the scheme for that database and add the user condition to
that scheme.

The only problems you might have is dealing with the same record being
changed by different users at different sites. But there are conflict
resolution bits for that. Alternatively you can add in stored procedures for
more control (which is what I'm currently working on)

Good luck it sounds like an interesting and challenging project.

Steve

"Georg Ledermann" <ledermann@gmx.de> wrote in message
news:3bb327a6_2@dnews...
> Hi,
>
> currently, I'm evaluating some replication tools
> for InterBase. I would like to choose IBReplicator
> if the following scenario can be handled with this
> tool:
>
> My customers often work at multiple locations: One
> is the central office, the others are subbranches
> located in other cities or laptop users. At every
> place there is installed my app with an InterBase
> server. The subbranches can connect to the central
> office by a dial up connection. The laptop users
> connect via a docking station to the central
> office OR one of the subbranches.
> Data can be inserted/changed/deleted on every
> place. Since everywhere the same data is needed,
> there is a need for replication: Every change of
> the data which took place at one location has to
> be replicated to all other places.
>
> My question is: Can this be done with
> IBReplicator? If yes: What, if some subbranches or
> laptop users are added? Does this affect the
> existing subbranches/laptop users?
>
> IMHO the main problem is the following thing: If
> there are more than two servers every server has
> to maintain which data has successfully replicated
> to every other server. IF this is right, every
> server must know the existence of every other
> server. But this is not good. It would be better
> if only the "main server" in the central office
> know how many server take part on replication.
>
> Additional question: Are global generators
> supported by IBReplicator?
>
> Best regards,
>   Georg
>