Real time adaptive monitoring

摘要

This disclosure describes systems and methods for dynamically and automatically adapting the level of data logging that occurs within a network of nodes, identifying causes of exceptions that occur within the network and resolving those causes to ensure that the network continues operating efficiently. A monitoring service may automatically increase/decrease data logging of various nodes within the network as it progresses through levels of a network searching for a node that is the source of a cause of an exception. Once identified, the monitoring service may process the logged data and either automatically resolve the cause of the exception or provide information to an operator for resolution.

主权项

1. A computer-implemented method for adaptive monitoring of a network having a plurality of connected processes, comprising:
monitoring the network for an occurrence of an event, wherein the event corresponds to an expected delivery time of a package to a destination; determining that the event exceeds a monitoring threshold, wherein the monitoring threshold relates to a probability as to whether the package will reach the destination by the expected delivery time; identifying a first process with an exception related to the event; automatically increasing a data logging for a plurality of processes connected to the first process; identifying a second process of the plurality of processes with the exception related to the event; automatically reducing the data logging of the plurality of processes other than the identified second process; determining that the second process is a source of the exception; determining a cause of the exception; correcting the cause of the exception; verifying that the exception is resolved; and automatically reducing the data logging for the second process.