On Nov 18, 2013, at 11:52 PM, Sanne Grinovero <sanne(a)infinispan.org> wrote:
> Neither the grouping API nor the AtomicMap work over hotrod.
> Between the grouping API and AtomicMap, I think the one that would make more sense
migrating is the grouping API.
> One way or the other, I think the hotrod protocol would require an enhancement - mind
raising a JIRA for that?
> For now I guess you can sacrifice performance and always sending the entire object
across on every update instead of only the deltas?
Doesn't Hot Rod even have support for deltas?
Our main concern is not so much performance but to implement the
correct isolation: ok we have no transactions, but still we'd be
messing with entries which weren't even supposed to be loaded, and
this data could get overwritten.
then why not just use a syntethic key for every key in the AtopmicMap?
e.g. if you have the following AtomicMap: (ak, {(k1, v1), (k2,v2)}) where:
ak - the key of the AtomicMap in the cache
k1, k2 - keys within the atomic map
then using some synthetic the keys: "ak_k1", "ak_k2" directly on the
remote cache would avoid the overwriting.
Cheers,
--
Mircea Markus
Infinispan lead (
www.infinispan.org)