Deploying a portion of a streaming application to one or more virtual machines

摘要

Profile data for streaming operators is collected and classified according to the function and relationship of operators, and split rules are constructed for the operators. A streams manager monitors performance of a streaming application, and when the performance needs to be improved, the streams manager determines from the split rules how to split the flow graph for the streaming application. The streams manager requests virtual machines from a cloud manager. In response, the cloud manager provisions one or more virtual machines in a cloud. The streams manager then modifies the flow graph so a portion of the flow graph is deployed to the one or more virtual machines in the cloud. In this manner a streaming application can dynamically evolve to increase its performance as needed.

主权项

1. An apparatus comprising:
at least one processor; a memory coupled to the at least one processor; a library of profile data collected from previous executions of a first plurality of operators in at least one streaming application, wherein the library of profile data classifies the first plurality of operators according to function; at least one split rule for at least one of the first plurality of streaming operators, wherein the at least one split rule is generated from information in the library of profile data and specifies at least one of:
a plurality of operators to keep together; anda plurality of operators to keep apart; a first streaming application residing in the memory and executed by the at least one processor, the first streaming application comprising a flow graph that includes a second plurality of operators that process a plurality of data tuples; and a streams manager residing in the memory and executed by the at least one processor, the streams manager monitoring performance of the first streaming application, and when performance of the first streaming application needs to be improved, the streams manager determines at least one applicable split rule for the second plurality of operators, determines from the at least one applicable split rule how to split the flow graph of the first streaming application to create a first portion to be deployed to at least one virtual machine, requests a cloud manager to provision the at least one virtual machine, and when the cloud manager provisions the at least one virtual machine, the streams manager modifies the flow graph to deploy the first portion to the at least one virtual machine.