Synchronizing "block"

[Guest]

I recently adopted a database when moving into a new clinic. It is used to
track our patients' lab results, treatments, etc. Four replicas exist, one of
which we use as the master. However, I am unsure if it is the Design Master
or not, though I doubt it is. I believe that exists somewhere offsite in
possession of the person in our organization who designed the database. At
the end of each clinic day, I synchronize all of the copies because patient
info has been entered in each replica in each of our different offices. Our
computers are not linked by internet, so I do this manually. I transfer
copies of each of the replicas onto our main computer, open our "master"
replica, then synchronize (from Tools>Replication>Synchronize) with each
replica (1 with 2, 1 with 3, 1 with 4, 1 with 2, 1 with 3) so that each
replica should have all of the updated information.

My problem is that the replicas are not synchronizing evenly. Specifically,
the master (and thus all other replicas) seems unable to accept new data from
one of the replicas (let's call it replica 2). However, all of the new data
from the other replicas is transfered to replica 2 during the
synchronization. The reason I first noticed this was taking place was that
the total record tally for replica 2 was increasing faster than the others.
Each time we add a new patient, we add a new record. After further
investigation, records of the new patients first added onto replica 2 were
not present in the other replicas, including the master, after
synchronization. However, new patients added to the other replicas were
present in replica 2 after synchronization.

I have no idea how to rectify this situation, or even where the problem lies.

Also, if anyone has any advice on a simpler method of synchronizing all of
these replicas without internet connection, I'd love some help with that.

 

[David W. Fenton]

I recently adopted a database when moving into a new clinic. It is
used to track our patients' lab results, treatments, etc. Four
replicas exist, one of which we use as the master. However, I am
unsure if it is the Design Master or not, though I doubt it is. I
believe that exists somewhere offsite in possession of the person
in our organization who designed the database. At the end of each
clinic day, I synchronize all of the copies because patient info
has been entered in each replica in each of our different offices.
Our computers are not linked by internet, so I do this manually. I
transfer copies of each of the replicas onto our main computer,

This is a terrible thing to do because it creates "dead replicas,"
by overwriting existing replicas and creating new ReplicaIDs that
then get lost when that replica is overwritten again. An explanation
is here (no word wrap and no spaces):

http://groups.google.com/group/comp.databases.ms-access/msg/9f5a84194
fa6c653

Continuing to do this is likely to cause replication errors that
will eventually reach a state where they simply cannot be corrected
(if you're not already there), and your data will be corrupted. Once
you can't be sure two replicas are identical after a synch, you're
already in a state that I'd consider unreliable enough to count as
corrupt.

open our "master"
replica, then synchronize (from Tools>Replication>Synchronize)
with each replica (1 with 2, 1 with 3, 1 with 4, 1 with 2, 1 with
3) so that each replica should have all of the updated
information.

You simply can't do it this way. Replicas have to by synched in
place. You need to connect the different machines to a LAN where the
partner replica for each is visible across the network. Or you need
to set up indirect or Internet replication, each of which enables
synchronization across a WAN.

But the synch can only happen with network-connected replicas. If
you can't have a network connection between the remote replicas and
one of the other replicas, then you just can't user Jet replication.

My problem is that the replicas are not synchronizing evenly.
Specifically, the master (and thus all other replicas) seems
unable to accept new data from one of the replicas (let's call it
replica 2). However, all of the new data from the other replicas
is transfered to replica 2 during the synchronization. The reason
I first noticed this was taking place was that the total record
tally for replica 2 was increasing faster than the others. Each
time we add a new patient, we add a new record. After further
investigation, records of the new patients first added onto
replica 2 were not present in the other replicas, including the
master, after synchronization. However, new patients added to the
other replicas were present in replica 2 after synchronization.

I have no idea how to rectify this situation, or even where the
problem lies.

I've explained what the problem is. Fixing it may not be simple. You
may have to figure out which records are not passing from Replica 2
to the others, and append them manually, then kill Replica 2 and
replace it with a new replica.

Also, if anyone has any advice on a simpler method of
synchronizing all of these replicas without internet connection,
I'd love some help with that.

If you can't have a network connection at least some of the time,
then you simply cannot use Jet replication.

It's really that simple.

And the problem you're having is exactly the first symptom of the
problems caused by doing it wrong.