"Bill Katelis" <bill@...> wrote in message
news:3D22A787.CB6D6E39@......

> "Jason Chapman (JAC2)" wrote:
>
> > Thoughts on high availability ticker DB's (that's what I call them):
> > Firstly for a 24 x 7 x 30.8 I would normally have 2 db's running in
> > parallel. That way you have additional fault tolerance and can run a
> > bespoke replication to allow inserts into either DB to propagate to the
> > other (this is so application specific, I've done it for one mfr co and
> > advised at least one other).
> >
>
> Jason - do you do this via a middleware type application which writes to

both

> db's ?
> Or do you use IBReplicator ?

Generally I would do this via replication at the database level. I have
always rolled my own replication, normally pretty basic and problem
specific. Never actually used the shrink wrapped replication in anger.
Simply we add an insert, update and delete trigger on the table we care
about to put a record into a to-do table that is specific for this function.

>
> >
> > This way if you need to you can take a DB down you can ensure it is
> > non-live, then let the replication deplete, then it is yours to do

whatever

> > with.
>
> Again - same question basically - How does the "replication deplete"?

The replication bot is always running and looking for records to replicate
from one DB to the other.

Jason Chapman
JAC2 Consultancy



Training - Development - Consultancy

Delphi, InterBase, Firebird, OOAD, Development lifecycle assistance,
Troubleshooting projects, QA.....

www: When I get round to it....

Mob: (+44) 07966 211 959 (preferred)

Tel: (+44) 01928 751088