Each node works on only a part of the data.
每个节点仅仅处理部分数据。
The current node works like an implicit parameter.
当前节点就像是一个隐式参数。
Replication on demand helps in the scenario where each node works with different pieces of data and will be combined at logical steps.
对于每个节点处理不同部分数据并在逻辑步骤中整合这些数据的情况,随需复制是很有帮助的。
应用推荐