Now in-memory caches use all active server nodes as affinity topology and it changes with each node join and exit. What differs from persistent caches behavior which uses BLAT (BaseLine Affinity Topology) as their affinity topology.
It causes problems:
- we lose (in general) co-location between different caches;
- we can't avoid PME when a non-BLAT node joins cluster;
- implementation should consider 2 different approaches to affinity
To handle these problems we should make in-memory and persistent cache work similar.