Hi, You actually collocate keys, not caches. Let me explain in more detail.
The primary node for a key is defined in two steps: 1. Map cache key to affinity key. This is defined by @AffinityKeyMapped annotation of AffinityKeyMapper. If none of them is provided, the affinity key is the same as cache key. 2. Map affinity key to node. This means that all cache keys that map to the same affinity key will be mapped to the same node, and this doesn't depend on the cache where these keys are saved (unless you explicitly provide different affinity configuration for them, of course). In your example, affinity key for Employee is organizationId. So all employees with the same organizationId will go to the same node. The organization that uses this ID as the cache key will also go to this node. If there is any other object (e.g. Company) that is stored in different cache, but identified by the same key, they will also happen to end up on the same node, but if Employee and Company are logically not related to each other, this is just a coincidence. Let me know if this makes sense to you. -Val -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/How-does-AffinityKey-mapped-tp6260p6285.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
