Analyzing distributed datasets
AssigneeAmazon Technologies, Inc.
InventorsMaarten Van Horenbeeck, Matthew Ryan Jezorek
Life of this patent
Estimating the term…
Abstract
Techniques for analyzing a dataset may be provided. For example, a configuration file may be accessed. The dataset may be analyzed based on a condition identified in the configuration file. A report may be generated and transmitted based on the analysis. Another report generated based on an analysis of another dataset according to another configuration file may be accessed. The dataset may be further analyzed based on this report to determine if a reported observation may also be associated with the dataset. If so, a confirmation may be generated and transmitted.
Claims20
A computer-implemented method, comprising:
- distributing, by a computer system, a configuration file to a plurality of nodes forming a point-to-point network, the plurality of nodes comprising a plurality of corresponding local datasets and a plurality of corresponding local agents, the plurality of corresponding local agents configured to analyze the plurality of corresponding local datasets based at least in part on the configuration file, the configuration file identifying data fields of a dataset to be analyzed according to a condition and an action to be performed if the condition is met, the computer system maintaining a privileged node of the plurality of nodes, the configuration file indicating that confirmations are to be protected with a public key associated with the privileged node and independently indicating whether reports are to be protected with the public key;
- receiving, by the computer system, a report from a node of the plurality of nodes, the report based at least in part on an analysis of the configuration file by a corresponding local agent of the node, the report indicating that analyzed data fields from a corresponding local dataset of the node meet the condition, the report comprising a summary of the analyzed data fields, the report received based at least in part on the action, the report protected using the public key in accordance with the indication in the configuration file;
- receiving, by the computer system, a confirmation from each of a set of nodes of the plurality of nodes, the confirmation indicating that a local dataset of the node comprises data associated with the summary, the confirmation received based at least in part on the action, individual confirmations protected using the public key in accordance with the indication in the configuration file such that non-privileged nodes of the plurality of nodes are unable to access protected content of the confirmation;
- comparing, by the computer system, a count of the set of nodes from which confirmations were received to a threshold;
- and
- determining, by the computer system, that the node is associated with a network-based issue based at least in part on the count being lower than the threshold.
One or more non-transitory computer-readable storage media storing computer-executable instructions that, when executed by a computing node configure the computing node to perform operations comprising:
- accessing a configuration file comprising a condition, the configuration file comprising an indication that reports are to be protected with a public key of a privileged computing node communicatively coupled with the computing node;
- determining that a portion of a local dataset meets the condition;
- generating a report based at least in part on meeting the condition, the report protected using the public key of the privileged computing node in accordance with the indication in the configuration file;
- accessing a second configuration file and a second report that is associated with a second computing node, the second report generated based at least in part on the second configuration file and comprising an observation associated with a second local dataset of the second computing node, the observation indicating that a second portion of the second local dataset meets a second condition from the second configuration file, the second configuration file comprising a second indication that messages are to be protected with the public key of the privileged computing node;
- analyzing the local dataset to determine whether the observation associated with the second local dataset is also associated with the local dataset;
- and
- generating a message indicative of an association between the local dataset and the observation based at least in part on the analyzing, the message protected using the public key of the privileged computing node in accordance with the second indication in the second report.
A system, comprising:
- a memory configured to store computer-executable instructions;
- and
- a processor configured to access the memory and execute the computer-executable instructions to collectively at least:
- transmit, to a first system and a second system, a configuration file comprising a condition, the configuration file indicating that confirmations are to be protected with a public key associated with the system and independently indicating whether reports are to be protected with the public key;
- receive, from the first system, a report comprising an observation indicative of the condition being met in a first dataset local to the first system, the report protected using the public key in accordance with the indication in the configuration file;
- receive, from the second system, a confirmation that the observation is associated with a second dataset local to the second system, the confirmation protected using the public key in accordance with the indication in the configuration file;
- and
- determine that the condition is associated with an issue based at least in part on the confirmation.
Dependencies are inferred from the claim text.