PREDICTED OUTPUTS IN A STREAMING ENVIRONMENT

摘要

According to embodiments of the disclosure, methods, systems, and computer program products for initializing a streaming application are disclosed. The method may include compiling code on a compiler system, the compiling of the code including establishing an operator graph having a plurality of processing elements including a first processing element and a second processing element. The compiling of code including receiving a first compiler directive specifying an operator sub-graph included in the operator graph, the operator sub-graph containing one or more processing elements from the plurality of processing elements. The compiling of code including receiving a second compiler directive identifying the first processing element, the first processing element configured to bypass the operator sub-graph by transmitting a predicted output tuple to the second processing element in response to determining that a processing condition exists in the operator graph.

主权项

1. A system for processing a stream of tuples, comprising:
a plurality of compute nodes, the plurality of compute nodes each having one or more processing elements to receive a stream of tuples, the processing elements operating on one or more computer processors, each processing element having one or more stream operators; and a stream manager, the stream manager configured to monitor the plurality of compute nodes and to determine whether a processing condition exists in the plurality of compute nodes; and a compiler, the compiler configured to receive a first compiler directive specifying one or more processing elements of the plurality of compute nodes to include in an operator sub-graph, the compiler configured to receive a second compiler directive identifying a first processing element of the one or more processing elements to bypass the operator sub-graph by transmitting one or more of the the predicted output tuples to a second processing element of the plurality of compute nodes, in response to determining that a processing condition exists in the plurality of compute nodes.