Accelerator Intermediary Nodes for Fine-Grained Write Control

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current storage services in provider networks often lack fine-grained control over I/O operations, particularly in multi-tenant environments, where customers may require varying levels of data durability, replication, and synchronization for different write and read requests, which existing technologies struggle to manage efficiently.

Innovation Solution

The implementation of accelerator intermediary nodes that allow for variable-granularity control settings requests, enabling customers to specify replication counts, data durability levels, transaction grouping, and synchronization modes on a per-write basis, using a data acceleration service as an intermediary between storage services and computing devices to optimize I/O operations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If multi-tenant virtualized storage nodes are used to increase resource utilization, then overall resource utilization improves, but fine-grained control over I/O operations for individual customers deteriorates

Engineering Contradiction:

Improveresource utilizationVSAvoidfine-grained control over I/O operations

Core Design Contradiction:

ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent segments control authority by introducing intermediary nodes that act as independent control points between storage nodes and customers. Each intermediary node can enforce different control settings (durability, replication, synchronization) for different customers, enabling fine-grained control within the multi-tenant environment without requiring separate physical storage infrastructure for each customer.

Inventive Principle:

Principle #1Segmentation

2Device complexity

If uniform replication and durability settings are applied to all write operations, then system simplicity is maintained, but customer-specific service quality requirements cannot be met

Engineering Contradiction:

Improvesystem simplicityVSAvoidcustomer-specific service quality

Core Design Contradiction:

Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent implements dynamic control settings where replication counts, durability levels, and synchronization modes can be adjusted on a per-write basis according to customer requirements. The intermediary nodes receive and enforce these variable control settings, allowing the system to adapt to different service quality needs without maintaining multiple static storage system configurations.

Inventive Principle:

Principle #15Dynamics

3Reliability

If write operations are propagated to multiple replicas synchronously to ensure data durability, then data durability improves, but write latency increases

Engineering Contradiction:

Improvedata durabilityVSAvoidwrite latency

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent implements partial action by allowing customers to specify different durability levels (e.g., requiring acknowledgment from 2 out of 3 replicas) rather than always requiring all replicas to be successfully written. The intermediary nodes enforce these partial durability requirements, enabling customers to trade off between durability and latency based on their specific needs for different data types.

Inventive Principle:

Principle #16Partial or excessive action

4Adaptability or versatility

If fine-grained control settings are implemented on a per-write basis, then service quality adaptability improves, but system complexity increases

Engineering Contradiction:

Improveservice quality controlVSAvoidsystem architecture

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent introduces intermediary nodes as mediators between customers and the underlying storage infrastructure. These intermediary nodes handle the complexity of enforcing fine-grained control settings (replication, durability, synchronization) without requiring changes to the core storage system architecture. The intermediaries translate customer requirements into appropriate storage operations, isolating complexity from both the storage infrastructure and customer applications.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10191663B1Using data store accelerator intermediary nodes and write control settings to identify write propagation nodes

Publication Date: 2019.01.29 AMAZON TECH INC

AI summary

An accelerator intermediary node (AIN) associated with a data store obtains an indication of a control setting to be applied with respect to a write request directed to a data item, where the control setting specifies a target for one or more of replication count, data durability, a transaction grouping with respect to a write request, or back-end synchronization node. Using the control setting, a write propagation node set is identified for the write request. The write propagation node set includes another accelerator intermediary node and/or a storage node of a data store. Respective operation requests corresponding to the write request are transmitted to one or more members of the write propagation node set. A write coordinator role may be verified prior to attempting a commit together of a plurality of write requests as part of a multi-write transaction.