Persistent Object Tracking in Dynamic Virtualized Environments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object tracking systems in virtualized and cloud computing environments fail to persistently track virtual machine instances across multiple starts, stops, and new instances due to the assignment of new IP addresses and host names, leading to difficulties in maintaining tracking across dynamic environments.
Innovation Solution
An object tracking system that assigns unique numerical identifiers (OIDs) and interval identifiers (IIDs) to objects, using a location broker model with an identity server and service location broker to track objects across restarts and clones, and implements a two-phase commit process to ensure accurate genealogy and resource management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a monitoring tool tracks objects using IP addresses and host names in virtualized environments, then it can identify objects during their operation, but it fails to maintain tracking across instance restarts and clones because new IP addresses and host names are assigned
Solution Approach 1:
The patent uses virtual disk copies and snapshots to track object identity. When a virtual machine is cloned or restarted, the monitoring system tracks the copy relationship through virtual disk identifiers and snapshot metadata, allowing continuous tracking without relying on transient IP addresses or host names that change with each instance renewal
Solution Approach 2:
The monitoring tool is integrated into the virtualization platform beforehand, enabling it to receive notifications about upcoming instance creation, cloning, or migration events. This preliminary awareness allows the system to prepare tracking identifiers in advance and maintain continuous object identity across lifecycle transitions
2Reliability
If a custom monitoring system is placed on top of the environment to detect object starts, stops, and clones, then object tracking can be maintained, but a bottleneck is introduced into the data flow
Solution Approach 1:
The monitoring functionality is merged directly into the virtualization platform's core components, such as the hypervisor or virtual machine manager. This integration allows the monitoring tool to access object lifecycle events and identity information directly from the platform's internal event streams and metadata structures, eliminating the need for separate monitoring layers that would create data flow bottlenecks
3Loss of information
If a monitoring system requires interface support from the environment, then it can obtain monitoring data, but its applicability is restricted to only those environments with such support
Solution Approach 1:
The monitoring solution is designed to work across multiple virtualization platforms and environments by implementing a universal interface layer that can adapt to different platform-specific event notifications and metadata formats. This allows the same monitoring tool to extract object identity and lifecycle information from various virtualization environments without requiring environment-specific modifications, thereby enhancing applicability while maintaining data access capability
Data Source
AI summary
Embodiments provide systems, methods, and computer program products for tracking objects within a domain and their possible changes and genealogy over time. An object identifier and an interval identifier are associated with an object that enters the domain with an unknown set of attributes and genealogy. The object identifier is analyzed to determine whether it is a pre-existing object identifier. A comparison of the interval identifier is performed to determine whether the object is a copy of a prior object. The object identifier is replaced with a new object identifier and the interval identifier reset when the object is new or a copy. Based on the object identifier and the interval identifier, which represents the possibility of change during each interval, information about the tracked object may be derived and analyzed to enhance performance.