Function Node Migration Using Call Chain Delay Data

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current function node migration methods in distributed systems rely on predictions, leading to potential discrepancies between predicted and actual resource usage, resulting in unnecessary migrations and suboptimal performance.

Innovation Solution

The method determines the to-be-migrated function node based on actual call chain data, specifically considering delay and interaction times between nodes to identify high-migration-degree intervals, ensuring that function nodes with high delays are migrated to reduce overall system delay and improve user experience.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Force

If function node migration is based on prediction, then migration can be performed in advance, but unnecessary migrations occur due to prediction errors

Engineering Contradiction:
Improveproactive migration capabilityVSAvoidmigration accuracy
Core Design Contradiction:
ForceVSReliability

Solution Approach 1:

The system implements feedback by continuously monitoring actual resource usage data from the distributed system and using this real information to validate and adjust migration decisions. The feedback loop compares predicted resource usage with actual usage, enabling the system to learn from prediction errors and improve migration accuracy over time while reducing unnecessary migrations.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs preliminary actions by conducting prediction-based migration assessments before actual migrations occur. By evaluating predicted resource usage thresholds in advance and cross-validating with real-time data, the system prepares migration candidates proactively while maintaining the option to cancel or adjust migrations based on actual conditions, thus avoiding unnecessary migrations.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If function nodes are migrated frequently to optimize resource allocation, then resource utilization improves, but system stability deteriorates due to excessive migrations

Engineering Contradiction:
Improveresource allocation efficiencyVSAvoidsystem stability
Core Design Contradiction:
ProductivityVSStability of the object's composition

Solution Approach 1:

The system applies dynamics by implementing adaptive migration thresholds that adjust based on system conditions and historical data. Rather than using fixed migration triggers, the system dynamically modifies migration criteria according to actual resource usage patterns, enabling flexible optimization of resource allocation while preventing excessive migrations that would compromise system stability.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system provides beforehand cushioning by establishing migration cooldown periods and stability buffers that prevent rapid successive migrations of the same function node. These protective measures cushion against excessive migration activity, ensuring that resource allocation optimization does not lead to system instability from frequent relocations.

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

3Speed

If migration decisions are made without verifying destination server compatibility, then migration speed increases, but migration failures occur when destination servers cannot run the function node

Engineering Contradiction:
Improvemigration decision speedVSAvoidmigration success rate
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The system performs preliminary compatibility verification between function nodes and destination servers before initiating migrations. By pre-checking server capabilities, resource availability, and compatibility requirements in advance, the system ensures that migration decisions are both fast and reliable, preventing migration failures while maintaining efficient deployment speed.

Inventive Principle:
Principle #10Preliminary action

4Reliability

If all function nodes are monitored equally for migration, then comprehensive coverage is achieved, but computational complexity increases

Engineering Contradiction:
Improvemigration decision comprehensivenessVSAvoidanalysis complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system applies local quality by implementing differentiated monitoring and analysis strategies for different function nodes based on their specific characteristics, criticality, and resource usage patterns. Rather than uniformly analyzing all nodes with the same computational intensity, the system applies targeted analysis depth appropriate to each node's local requirements, reducing overall computational complexity while maintaining comprehensive migration decision-making.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP4068727B1Method for migrating function node, and related device
Publication Date: 2024.07.31 HUAWEI CLOUD COMPUTING TECHNOLOGIES CO LTD
  • EP4068727B1 patent drawingFigure 1~2
  • EP4068727B1 patent drawingFigure 3~4
  • EP4068727B1 patent drawingFigure 5~6

AI summary

This application provides a function node migration method and a related device, where the method includes: obtaining access related information of at least one call chain, where the access related information includes at least one of the following information: a delay of each of N intervals included in the at least one call chain, and a quantity of interaction times between function nodes in each of the N intervals, where N is a positive integer greater than or equal to 1; determining, based on the access related information, P key nodes from a plurality of function nodes that form the N intervals; and migrating the P key nodes. In the foregoing technical solution, a to-be-migrated function node is determined based on actual data of the call chain. Therefore, the to-be-migrated function node determined in the foregoing technical solution meets a migration requirement. Therefore, in the foregoing technical solution, the to-be-migrated function node can be determined based on an actual requirement, which reduces unnecessary migration.