Event-Driven Computational Storage for Near-Data Function Execution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current orchestration solutions for edge computing face challenges such as vendor-lock, sub-optimal performance due to incorrect QoS requests, and difficulty in selecting resource allocations for microservices, leading to overprovisioning and higher costs, especially in heterogeneous hardware environments.
Innovation Solution
Implementing an intent-driven model that maps service level objectives across systems and resources using nested and graduated Service Level Agreements (SLAs) to achieve quality of service, allowing abstraction from detailed resource definitions and enabling self-adaptive orchestration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional orchestration solutions are used for edge computing, then vendor-specific implementations can be deployed, but vendor-lock occurs and adaptability to different hardware environments is reduced
Solution Approach 1:
The patent introduces an intermediary layer between the orchestration system and heterogeneous hardware resources. This intermediary translates high-level service level objectives into hardware-specific configurations, enabling vendor-neutral orchestration while maintaining adaptability to different hardware environments through standardized interfaces that map to underlying diverse resources.
Solution Approach 2:
The orchestration system implements universal resource allocation mechanisms that can handle multiple types of hardware platforms and service scenarios through a unified framework. The system provides multi-functional capabilities to manage diverse compute, storage, and network resources across different vendors using common orchestration primitives and standardized service level agreements.
2Adaptability or versatility
If detailed resource definitions are specified for QoS requests, then resource allocation can be precise, but the system becomes less adaptable and requires more complex configuration
Solution Approach 1:
The system implements dynamic resource allocation where QoS parameters are not fixed but can be adjusted based on runtime conditions and service level objectives. The orchestration system dynamically translates high-level service requirements into appropriate resource allocations, allowing flexibility while maintaining QoS guarantees through adaptive resource management that responds to changing conditions.
Solution Approach 2:
The patent employs parameter changes by allowing service level objectives to be defined at multiple levels of abstraction. The system transforms high-level service level objectives into specific resource allocation parameters dynamically, enabling the same service objective to be satisfied through different resource configurations depending on available capacity and hardware characteristics.
3Reliability
If resource allocations are overprovisioned to ensure QoS, then service reliability is improved, but resource utilization efficiency decreases and costs increase
Solution Approach 1:
The orchestration system implements feedback mechanisms that monitor actual resource usage and service level achievement in real-time. Based on this feedback, the system dynamically adjusts resource allocations to maintain QoS guarantees while optimizing utilization. The feedback loop enables the system to scale resources up or down based on actual demand patterns rather than static overprovisioning.
Solution Approach 2:
The system applies partial provisioning by allocating resources based on actual service level objectives rather than maximum possible capacity. Instead of overprovisioning for all scenarios, the system provisions exactly the amount needed to meet service level targets, using graduated service level agreements that allow different levels of resource allocation for different service priorities and performance requirements.
Data Source
AI summary
Various systems and methods for implementing computational storage are described herein. An orchestrator system is configured to: receive, at the orchestrator system, a registration package, the registration package including function code, a logical location of input data for the function code, and an event trigger for the function code, the event trigger set to trigger in response to when the input data is modified; interface with a storage service, the storage service to monitor the logical location of the input data and notify a location service when the input data is modified; interface with the location service to obtain a physical location of the input data, the location service to resolve the physical location from the logical location of the input data; and configure the function code to execute near the input data.