Endpoint Profiling via Single-Response Scan and Data Labeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional endpoint scanning and profile generation processes are time-consuming, resource-intensive, and prone to errors due to the need for multiple discrete data requests and the potential for mismatched data types, leading to inefficient and crash-prone operations.
Innovation Solution
A structured request is used to gather endpoint parameters with anticipated data types, allowing for a single response that includes metadata, which is stored in a data lake for organized processing and labeling, enabling efficient generation of a profile report.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple discrete data requests are used to gather endpoint parameters, then comprehensive endpoint profiling is achieved, but scan time increases to tens of minutes and resource consumption increases
Solution Approach 1:
The patent combines multiple discrete data requests into a single scan message that can be processed in one operation. The structured request format allows multiple inquiries about different endpoint parameters to be merged into one communication, reducing scan time from tens of minutes to one or two minutes while maintaining comprehensive profiling capability.
Solution Approach 2:
The scan message structure is designed to be universal, capable of handling multiple types of endpoint parameters simultaneously. The system can accommodate various data types (strings, numbers, booleans, arrays, objects) within a single request framework, eliminating the need for separate discrete requests for different parameter types.
2Measurement precision
If multiple discrete data requests are used to gather endpoint parameters, then comprehensive endpoint profiling is achieved, but resource consumption increases
Solution Approach 1:
The patent merges multiple resource-intensive discrete requests into a single scan message operation. By consolidating multiple data gathering operations into one structured request, the system significantly reduces CPU usage, memory allocation, and network bandwidth consumption while maintaining the ability to collect comprehensive endpoint information.
3Productivity
If conventional scanning processes are used, then endpoint data is collected, but error rates increase due to mismatched data types and crash-prone operations
Solution Approach 1:
The patent applies preliminary action by defining anticipated data types for each endpoint parameter in advance within the structured request. This allows the system to prepare appropriate handling logic before receiving the actual data, preventing type mismatches and errors. The response parser uses these pre-defined expectations to correctly interpret returned data, significantly reducing error rates.
4Adaptability or versatility
If systems and applications change on endpoints, then updated information is available, but conventional requests must be correspondingly updated or errors occur
Solution Approach 1:
The patent implements dynamics by making the structured request format adaptable to changing endpoint configurations. The system can dynamically adjust to new systems and applications on endpoints without requiring manual updates to request structures, as the response parser handles various data types and formats flexibly based on pre-defined anticipated types rather than rigid fixed schemas.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A method of profiling an endpoint includes generating a structured request including a set of inquiries, each of which being directed to an endpoint parameter and including a parameter name. The method includes generating a scan message including the set of inquiries and an additional inquiry directed to an additional endpoint parameter. The method includes receiving a single response from the endpoint including raw parameter data responsive to the set of inquiries and the additional inquiry. The method includes storing the raw parameter data in a data lake organized according to extracted metadata. The method includes labeling each data portion using the parameter name and the extracted metadata. The method may include accessing, from the data lake, portions of the raw parameter data responsive to the set of inquiries based on labels associated therewith. The method includes generating a profile report derived from the accessed data.