Indirect Event Stream Correlation Using Uniqueness Tokens
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed systems face challenges in correlating event streams across multiple computers due to clock synchronization issues and the need for modifications in existing components for logging-based correlation.
Innovation Solution
An indirect event stream correlation process that uses task identifiers and ordering identifiers unique to each computer, along with a uniqueness token, to establish a direct correlation between events on different computers, allowing for the assembly of a complete list of events across computers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If time-based correlation is used to correlate event streams across multiple computers, then event ordering can be established, but clock synchronization issues cause incorrect temporal ordering
Solution Approach 1:
The patent introduces a uniqueness token as an intermediary mechanism that mediates the correlation between events on different computers. Instead of directly comparing timestamps from potentially unsynchronized clocks, the system uses these tokens as a trusted intermediary that establishes correct temporal relationships. The token is passed between computers during communication, allowing event streams to be correlated reliably without requiring clock synchronization.
2Loss of information
If logging details are placed in all calls between components to enable correlation, then event ordering can be reconstructed, but modifications to existing components are required
Solution Approach 1:
The patent applies preliminary action by pre-generating uniqueness tokens before events occur and embedding them in communication between components. Rather than requiring post-hoc modifications to log detailed information, the correlation information is prepared in advance and attached to events as they are generated, eliminating the need for extensive retroactive modifications to existing components.
3Measurement precision
If unique task identifiers and ordering identifiers are used for each computer, then events can be correlated within the same computer, but direct correlation between events on different computers is lost
Solution Approach 1:
The patent implements the nested doll principle by embedding the uniqueness token within the existing task identifier and ordering identifier structure. The token is nested inside the event data that already contains computer-specific identifiers, allowing both levels of correlation to coexist: the outer structure maintains computer-specific identification while the inner nested token provides cross-computer correlation information.
Data Source
AI summary
Various technologies and techniques are disclosed that indirectly correlate event streams on multiple computers. Event details are retrieved for multiple data streams from each of the computers. The event details include a task identifier and ordering identifier for each event in the data streams on each computer. The task identifiers and ordering identifiers are unique to each computer, allowing the events to be correlated with the other events on the same computer only. A uniqueness token is retrieved that provides direct correlation information about how a particular event on the first computer is directly correlated to a particular event on the second computer. An indirect event stream correlation process uses the directly correlated information in combination with the respective task identifiers and respective ordering identifiers from the first and second computers to establish an ordering between the events on the first computer and the events on the second computer.