Oh yeah, I meant to mention local replication too but forgot. What kind of Region-level API are you thinking of?
Something that the new cluster could call via a pool to pull the data for a region? -- Mike Stolz Principal Engineer, GemFire Product Manager Mobile: 631-835-4771 On Fri, Dec 23, 2016 at 12:48 PM, Anthony Baker <aba...@pivotal.io> wrote: > I don’t know of anyway current API that behaves the way you want. > Snapshot imports (`gfsh import data`) avoid firing listeners AND wan > replication. Getting that working correctly was pretty hairy :-) > > I think you would want to avoid local replication costs associated with a > put() as well. Perhaps a Region-level API would make more sense. > > Anthony > > > On Dec 23, 2016, at 9:29 AM, Michael Stolz <mst...@pivotal.io> wrote: > > > > One way to get a new site populated with data would be through a WAN > > gateway by doing a put(k, get(k)) on every local entry in a Partitioned > > Region via a server side function. That has the unfortunate side-effect > of > > firing CacheWriters, CacheListeners and AsyncEventListeners though. > > > > Is there a mechanism by which we could force all data in a Region to be > > pushed across the WAN gateway without invoking CacheWriters, > CacheListeners > > and AsyncEventListeners? > > > > This would make cloud-bursting scenarios work really well. > > > > It would also help anybody who is trying to bring up an additional site > or > > additional cluster for any production workload. > > > > -- > > Mike Stolz > > Principal Engineer, GemFire Product Manager > > Mobile: 631-835-4771 > >
