Using pagerank algorithm-based lock analysis to identify key processes for improving computing system efficiency

摘要

A system is described for identifying key lock contention issues in computing devices. A computing device is executed and lock contention information relating to operations during execution of the computing device is recorded. The data is parsed and analyzed to determine blocking relationships between operations due to lock contention. Algorithms are implemented to analyze dependencies between operations based on the data and to identify key areas of optimization for performance improvement. Algorithms can be based on the Hyperlink-Induced Topic Search algorithm or the PageRank algorithm.

主权项

1. A method comprising:
executing a computing device for a predetermined time interval and recording data related to the execution of the computing device into a log; analyzing the data recorded into the log to identify blocking relationships between processes executing on the computing device, wherein blocking relationships comprise lock contention information related to a first process being blocked by a second process due to the first process being unable to acquire a lock on an object because the second process is holding the lock on the object; and plotting the data into a graph data structure comprising:
a plurality of nodes, each node representing one of the processes executing on the computing device; andlinks between the nodes, wherein a link between two nodes describes a blocking relationship between the two nodes, including indication of which of the two nodes blocks the other; setting a predetermined initial optimization priority score for each node; performing a first update of the optimization priority score of each node by calculating, for the each node, a first updated optimization score based on links between the each node and nodes blocked by the each node and on the initial optimization priority scores of the nodes blocked by the each node; performing a second update of the optimization priority score of each node by calculating, for each node, a second updated optimization score based on links between the each node and nodes blocked by the each node and on the first optimization priority scores of the nodes blocked by the each node; performing subsequent updates of the optimization priority score of each node until predetermined criteria for terminating the process are met; wherein the optimization priority score of a given node is a value that is deemed to correspond to an extent of impact of the given node on efficiency of operation of the computing device; and indicating one or more nodes on a visual display, wherein the indicating the one or more nodes further relates to how the indicated one or more nodes is deemed to impact efficiency of operation of the computing device based on a most recently updated optimization priority score of the indicated one or more nodes.