Shuffle-Sharded Load Balancing for Cloud Service DDoS Isolation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional cloud-hosting infrastructure systems lack configurable flexibility and telemetry information for layer 3 and 4 routing, leading to inefficiencies in resource allocation and increased vulnerability to DDOS attacks, which can impact other services.

Innovation Solution

Implementing shuffle sharding with a load-balancer program that routes data traffic through ingress host cells based on mapping data, captures telemetry information, and isolates services using quarantine cells to mitigate DDOS attacks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If conventional layer 3 network load balancing functions are used, then routing operations can be performed, but telemetry information related to routing actions is not accessible to the service provider

Engineering Contradiction:
Improvetelemetry informationVSAvoidnetwork architecture complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent introduces a layer 7 load balancer as an intermediary component between the layer 3 network load balancer and the service provider's computing resources. This layer 7 load balancer acts as a mediator that receives telemetry information from the layer 3 load balancer and makes it accessible to the service provider, thereby resolving the information loss without significantly increasing architectural complexity

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent adds a layer 7 dimension to the existing layer 3 networking architecture. By introducing application-layer load balancing capabilities, the system gains access to telemetry information that was previously unavailable at the network layer, effectively adding another dimension of observability and control

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Reliability

If conventional security services drop traffic exceeding threshold amounts, then DDOS attack mitigation can be performed, but acceptable data traffic is also dropped indiscriminately

Engineering Contradiction:
ImproveDDOS attack mitigationVSAvoiddata traffic throughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies different security policies to different sources of traffic based on their characteristics. Instead of a uniform threshold-based approach, the system identifies and treats malicious traffic differently from legitimate traffic, allowing acceptable data traffic to pass through while blocking only the harmful portions

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system uses telemetry information to monitor traffic patterns and adjust security responses dynamically. By analyzing traffic characteristics in real-time, the security services can distinguish between DDOS attacks and legitimate high-volume traffic, adjusting drop thresholds based on feedback from observed traffic patterns

Inventive Principle:
Principle #23Feedback

3Reliability

If conventional security services drop suspicious traffic indiscriminately, then security can be maintained, but noisy neighbor problems occur impacting other computing services

Engineering Contradiction:
ImprovesecurityVSAvoidnoisy neighbor problems
Core Design Contradiction:
ReliabilityVSObject-generated harmful factors

Solution Approach 1:

The patent segments the traffic handling architecture into isolated components, including separate load balancers and computing resource groups. This segmentation contains DDOS attacks and malicious traffic to specific segments, preventing them from affecting other computing services hosted on the same infrastructure

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent extracts malicious traffic from the general traffic flow and routes it through separate security handling paths. By taking out suspicious traffic early in the processing chain and treating it differently from legitimate traffic, the system maintains security while preventing noisy neighbor effects

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS20250385868A1Load balancer and shuffle sharding for cloud-hosted services
Publication Date: 2025.12.18 STRIPE LLC
  • US20250385868A1 patent drawing
  • US20250385868A1 patent drawing
  • US20250385868A1 patent drawing

AI summary

Embodiments include hardware and software resources of a distributed computing system for routing user data traffic to computing resources organized using a shuffle sharding arrangement. Layer 3 (L3) network load balancers proxy or route user data traffic requests to layer 4 (L4) transport load balancers. A L4 transport load balancer proxies and routes the requests to certain ingress cells that are assigned or mapped to the hosted computing services of the user requests according to proxy or routing functions and mapping data. The assignments between cells and computing services may be implemented by a cell manager program when onboarding the computing services in accordance with the shuffle sharding arrangement and configuration. The transport load-balancer may impose and enforce the shuffle sharding by routing user data to ingress cells assigned to the computing services using previously determined mappings data stored in a mappings database (or data file).