(dropping users@ - try not to cross post please) On Thu, Oct 31, 2013 at 12:13:18PM -0500, Alex Ough wrote: > All, > > While I'm waiting for the permissions, I'd like to bring up a discussion on > the overall architecture to support this feature. > > There can be 2 different approaches as below. > 1. master - slave architecture : the manual changes are allowed only in one > master management server, and those in other servers are either prohibited > or discarded. > 2. multiple source architecture : all management servers allow manual > changes and any change in any server will be propagated to the rest of > servers. > > I'm not sure if we can impose the restriction of #1. > If not, #2 is a little more complicated, but the only issue is to > distinguish the events from either manual jobs or triggered automatic > processing, which I'm currently working on. > > Your comments/recommendations will be very appreciated. > Thanks > Alex Ough
Marcus - you might want to track this discussion, since I know that you were surprised that there wasn't a native sync between regions as part of the initial region feature. -chip
