Distributed Timer Replication in Overlay Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer systems face challenges in ensuring the availability of timers across a network, particularly when nodes crash or malfunction, leading to potential data loss and difficulties in resuming processes due to the lack of guaranteed timer activation at specified wake times.
Innovation Solution
The implementation of distributed timers in an overlay network, where timers are replicated across multiple nodes for durable storage, ensuring that if one node fails, others can take over and activate processes at the designated time, maintaining data availability and process continuity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If timers are stored in durable storage at a single node, then storage reliability is improved, but availability is worsened because the node may be down at the specified wake time
Solution Approach 1:
The patent replicates timer data across multiple nodes in the distributed network. Each node maintains a copy of the timer information in its durable storage, ensuring that if one node is unavailable, other nodes can still activate the process at the specified wake time. This copying mechanism directly resolves the contradiction by maintaining storage reliability while improving timer availability through redundancy.
Solution Approach 2:
The timer management functionality is segmented across multiple independent nodes rather than centralized at a single node. Each node independently stores and can execute timer activation, distributing the responsibility for timer availability. This segmentation allows the system to maintain reliability even when individual nodes fail, as other segments (nodes) can take over.
2Reliability
If machines are clustered to provide increased availability for timers, then timer availability is improved, but network traffic is worsened due to re-distribution of timers when nodes are added or removed
Solution Approach 1:
The patent assigns specific identifier ranges to specific nodes, creating a stable mapping between timers and nodes. Timers with identifiers within a node's assigned range are stored locally at that node without requiring redistribution. This local quality approach reduces network traffic by eliminating the need to re-distribute timers during node additions or removals, while maintaining timer availability through the overlay network structure.
3Device complexity
If a single node is responsible for timer activation, then device complexity is reduced, but reliability is worsened due to single point of failure
Solution Approach 1:
Each node in the overlay network autonomously manages timers within its assigned identifier range without requiring coordination or communication with other nodes for timer activation. This self-service capability allows any node to independently activate processes for timers in its range, eliminating single points of failure while maintaining relatively simple individual node complexity. The distributed autonomous operation improves reliability without significantly increasing overall system complexity.
Data Source
AI summary
The present invention extends to methods, systems, and computer program products for using distributed timers in an overlay network. Embodiments of the invention can be used to replicate timers within an overlay network and can be used to make progress towards completing a portion of work within the overlay network based on a replicated timer. Accordingly, embodiments of the invention can be used to increase the availability of timer data as well as compensate for node failures within an overlay network.