Now that EasySync is close to its official "v1" launch, I decided to switch gears a little. Late this afternoon I spent some time working on an idea that Bill Heizer of FileMaker Inc. had proposed: Use the EasySync engine to create a server-to-server sync engine. The movie below shows the results.
In the movie, the "publisher" database is pushing out updates to the "subscriber" database. The subscriber database is essentially an exact copy of the the publisher database in terms of the data that is stored in it. As the movie shows, both binary and standard data are supported, as are deletions. The development work is being done with an Amazon EC2 server (the server that the "publisher" database is hosted on) and a 4-year old Mac Mini (on which the "subscriber" database is hosted). So there's nothing special here in terms of the hardware being used.
I expected to run into all sorts of challenges with this, but pulling in pieces of the EasySync engine made this relatively easy. The only real "gotcha" that I ran into is that Perform Script on Server isn't supported in server-side scripts. Go figure!
This early testing is very promising, but there is still a lot to be done: testing, documentation to be written, and so on. Regardless, if all goes as planned, then ServerSync will soon be available as an open source database replication solution.