Can I set up MySQL as a slave only replicating structure changes? (
ALTER TABLE etc.). I've got 2 reasons for this:
- Multiple developers with development branches in code which should always work with a 'fresh' datastructure, fetched from a 'main' development database.
- Multiple projects / clients which share (part of) the same code/scripts (think of a symlinked 'core' directory), but with different databases (1 per client), for which structure should be updated simultaneously with code changes in the shared portion of the code, with a minimum amount of downtime.
If this isn't possible, what are other good ways to sync mysql-datastructures without the actual data across sites / databases / servers? And no, structional changes don't happen all that often, but with enough different databases hanging around it's quite some time spend syncing all instances, and I'd like to both limit downtime and spend my time on more interesting jobs ;).