Lock Manager Migration for Network Latency Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-node computer networks, accessing shared resources can lead to increased network traffic and latency due to the burden on compute nodes, particularly when lock managers are not optimally located relative to nodes seeking resource access.
Innovation Solution
The method involves migrating lock manager instances to coupling facility locations closest to nodes that need to access shared resources, thereby reducing network traffic and latency by offloading lock management tasks and utilizing network-attached devices to equidistant locations based on access latency metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If lock manager instances are located on compute nodes, then resource access control is maintained, but network traffic and latency increase due to distant access
Solution Approach 1:
The lock manager instance is made dynamic by migrating it from a static location on compute nodes to a mobile location on network-attached devices. The system continuously monitors access latency metrics and relocates lock manager instances to optimal positions, transforming a static architecture into a dynamic one that adapts to changing network conditions and access patterns.
Solution Approach 2:
Network-attached devices serve as intermediaries between compute nodes and lock manager functionality. By placing lock manager instances on these intermediary devices rather than directly on compute nodes, the system reduces access latency while maintaining the separation of compute and storage/resources management functions.
2Loss of energy
If lock manager instances are migrated to network-attached devices, then network traffic is reduced, but system complexity increases
Solution Approach 1:
The system implements self-service through automated monitoring and migration of lock manager instances. The network-attached devices autonomously monitor access latency metrics and perform migrations based on predefined criteria, eliminating the need for manual intervention and reducing the operational complexity despite the increased architectural complexity.
Solution Approach 2:
The system changes the location parameter of lock manager instances from compute nodes to network-attached devices based on monitored access latency parameters. This parameter-driven approach allows the system to optimize network traffic by relocating instances to positions that minimize access distance, while the automation of this process manages the complexity.
3Speed
If lock manager instances are placed closer to accessing nodes, then access latency decreases, but network utilization becomes unbalanced
Solution Approach 1:
The system applies local quality by placing lock manager instances on network-attached devices that are locally accessible to specific groups of compute nodes. Rather than centralizing all lock management in one location, the system creates multiple local lock manager instances positioned near different node groups, optimizing access speed for each local region while balancing overall network utilization.
Data Source
AI summary
A method and system are disclosed for migrating network resources to improve network utilization, for use in a multi-node network wherein nodes of the network share network resources. The method comprises the steps of identifying a group of nodes that share one of the network resources, and identifying one of the nodes satisfying a specified condition based on at least one defined access latency metric. The shared resource is moved to the identified one of the nodes to reduce overall access latency to access the shared resource by said group of nodes. One embodiment of the invention provides a method and system to synchronize tasks in a distributed computation using network attached devices (NADs). A second embodiment of the invention provides a method and system to reduce lock latency and network traffic by migrating lock managers to coupling facility locations closest to nodes seeking resource access.