I found the document here
https://apacheignite-sql.readme.io/docs/select *SELECT queries can be executed against both replicated and partitioned data. When executed against fully replicated data, Ignite will send a query it to a single cluster node and run it over the local data there. On the other hand, if a query is executed over partitioned data, then the execution flow will be the following: The query will be parsed and split into multiple map queries and a single reduce query. All the map queries are executed on all the nodes where required data resides. All the nodes provide result sets of local execution to the query initiator (reducer) that, in turn, will accomplish the reduce phase by properly merging provided result sets.* Seems for replicated data, it always executes on a single node. then in this case is there a way to speed up the query by let all the nodes runs in parellel? -- Sent from: http://apache-ignite-users.70518.x6.nabble.com/