Federated Cluster Pod Creation with Globally Unique Instance IDs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In federated clusters, instance IDs assigned by each controller start from 0, leading to pod instances in different containerized application clusters having the same ID, making it impossible to control the deployment or running of stateful applications in a defined order.
Innovation Solution
A method and apparatus for creating pod instances in a federated cluster by configuring unique instance identifiers for controllers of different sub-clusters based on instance requirement and allocation numbers, ensuring distinct IDs across sub-clusters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If each controller assigns instance IDs starting from 0 in federated clusters, then the deployment process is simple and straightforward, but pod instances in different sub-clusters have duplicate instance IDs, making it impossible to control stateful application deployment in a defined order
Solution Approach 1:
The patent segments the global instance ID space into separate ranges for different sub-clusters. Each controller is assigned a specific segment of instance IDs based on its sub-cluster identity, ensuring that instance IDs are unique across the federated cluster while maintaining simple local assignment logic within each sub-cluster.
Solution Approach 2:
The patent implements local quality by allowing each sub-cluster to have its own instance ID assignment characteristics (starting from 0 locally) while the global system ensures uniqueness through segmented allocation. This maintains the simplicity of local deployment processes while achieving global instance ID uniqueness.
2Reliability
If instance identifiers are configured globally across federated clusters, then pod instances have unique IDs enabling ordered deployment, but the system complexity increases due to coordinated instance allocation across multiple controllers
Solution Approach 1:
The patent applies preliminary action by pre-configuring instance ID ranges for each controller based on their sub-cluster assignments before deployment begins. This preliminary segmentation eliminates the need for complex real-time coordination during deployment, as each controller independently assigns IDs from its pre-assigned range.
Solution Approach 2:
The patent segments the instance ID management responsibility across multiple controllers, with each controller managing a specific segment. This segmentation reduces the coordination complexity by allowing independent operation within each segment while ensuring global uniqueness through the segmented structure.
3Productivity
If stateful applications are deployed in federated clusters with duplicate instance IDs, then resource utilization across multiple sub-clusters is maximized, but the ability to control deployment order and manage application dependencies is lost
Solution Approach 1:
The patent introduces a new dimension to instance ID management by incorporating sub-cluster identification into the instance ID structure. This dimensional change allows the system to maintain unique instance IDs across federated clusters while preserving the ability to control deployment order, thereby enabling both high resource utilization and manageable deployment operations.
Data Source
AI summary
A method and an apparatus for creating pod instances in a federated cluster includes: acquiring application configuration information of a target application, and determining an instance requirement number of the target application according to the application configuration information; selecting at least two sub-clusters from sub-clusters included in the federated cluster, and determining an instance allocation number of each of the at least two sub-clusters according to the instance requirement number; creating a controller for each of the at least two sub-clusters, and configuring an instance identifier for a controller of a corresponding sub-cluster according to the instance requirement number and the instance allocation number of each of the at least two sub-clusters, where instance identifiers configured for controllers of different sub-clusters are different; and creating a pod instance in the corresponding sub-cluster by using the controller according to configured instance identifiers.


