well... I guess, I could use some generic class for all entities, so there is only one class - *Item*, with "type" and "userId" attributes on each node, ex:
- *User*("id":"123","name":"john",age:99) -> *Item* ("type":"user", "userId":"123", "name":"john",age:99) - *Purchase*("id":"2222","userId":"123","price":100) -> *Item* ("type":"purchase", "purchaseId":"2222", "userId":"123", "price":100) - *Account*("id":"11111","userId":"123","balance":10000) -> *Item* ("type":"account", "accountId":"11111", "userId":"123","balance":10000) - etc. in that case, I hope, it would be possible to implement custom sharding strategy - based on attribute "userId" (all Items related to userId will be placed in the same partition - user, purchases of that user, accounts of that user, etc - that is what I'm looking for). But following this approach I lose object-oriented features of OrientDb API and custom types definition via schema, maybe I will loose something else too, which I don't know about yet :( maybe there is a better way to do such thing? -- --- You received this message because you are subscribed to the Google Groups "OrientDB" group. To unsubscribe from this group and stop receiving emails from it, send an email to orient-database+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.