@
Info@MsAccess.nl
interesting concept - I can see that working between two databases, but say you have one 'master' and 30 'satellite' databases. How does it work then?
Each satellite creates a text file (presumably one for each table or perhaps a single xml file) and by whatever method sends it to the master who then imports and deletes or archives the file? But what about the other way? The master has read the data from satellite A and then needs to tell all the other satellites about the updates - does it create 30 text files, one for each satellite? or a single text file which all the other satellites can read - but at what point can that file be deleted or archived?
Or does the satellite produce some sort of return to the master that says 'processed that file' (plus no doubt an internal flag to prevent the satellite trying to upload the file again) and when all satellites have done the return the master can delete/archive the file?
You also have the potential for data clashes - satellite A and satellite B both modify some customer detail for the same customer - both send up up a text file and both are imported to the master - but which one is right? How do you manage that?
Similarly, two satellites create a new record for the same customer at the same time - both get uploaded. They have discreet ID's so not problem there, but you now have the same customer in the system twice. Made doubly difficult if there is a small typo difference between say customer name - in a project I was involved in a few years ago, we discovered McDonalds had been entered into the system 21 times, each with a slightly different spelling. The correct one is McDonald's but we also had McDonalds, Macdonalds, Mc Donalds, to name just a few.
Appreciate it very much depends on what the app is actually doing but would be interested to know if you are having to cope with the above scenarios and if so, how.