Ad Network Request Rejection for Overload Mitigation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Real-time bidding systems in online advertising networks face performance degradation due to system overload, leading to increased failure rates and inefficiencies in matching advertisers with advertisement opportunities, as they struggle to handle a large volume of requests effectively.

Innovation Solution

Implementing a system that selectively rejects incoming data access requests when system overload is detected, using a performance tracking module to predict execution time delays and decide whether to fulfill or reject requests, thereby preventing catastrophic failures and maintaining system performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the system attempts to reach more advertisers or advertisement agencies, then the coverage and matching capability improve, but the network speed slows down and the failure rate increases

Engineering Contradiction:

ImprovecoverageVSAvoidfailure rate

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system performs preliminary actions by maintaining a curated list of prioritized advertisers and advertisement agencies that are pre-approved for access. This allows the system to quickly determine which incoming requests should be fulfilled without attempting to process every possible requester, thereby maintaining high coverage of valuable advertisers while avoiding overload from excessive requests

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system applies local quality by treating different advertisers and agencies differently based on their priority status. High-priority entities receive preferential treatment with guaranteed response attempts, while lower-priority entities are subject to rate limiting and selective rejection. This differentiated approach ensures reliable service for critical advertisers while managing overall system load

Inventive Principle:

Principle #3Local quality

2Adaptability or versatility

If the system attempts to reach more advertisers or advertisement agencies, then the coverage and matching capability improve, but the network speed slows down

Engineering Contradiction:

ImprovecoverageVSAvoidnetwork speed

Core Design Contradiction:

Adaptability or versatilityVSSpeed

Solution Approach 1:

The system pre-establishes priority relationships with advertisers and agencies before processing requests. This preliminary classification allows for O(1) decision-making on request fulfillment without complex real-time evaluations, maintaining high network speed while preserving comprehensive coverage of prioritized advertisers

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system segments advertisers and agencies into priority tiers, with each tier receiving different levels of service. This segmentation allows the system to maintain fast response times for high-priority entities while applying rate limiting to lower-priority entities, thereby preserving overall network speed while maintaining broad coverage

Inventive Principle:

Principle #1Segmentation

3Adaptability or versatility

If the system fulfills all incoming requests, then the service completeness improves, but the system becomes overloaded and performance degrades

Engineering Contradiction:

Improveservice completenessVSAvoidperformance

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The system deliberately chooses partial action by selectively fulfilling only high-priority requests while rejecting or rate-limiting lower-priority ones. This partial fulfillment strategy prevents system overload and maintains performance for critical advertisers, accepting that not all requests can be served equally

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The system dynamically changes the parameter of request acceptance based on current system load and request priority. When the system is overloaded, it adjusts its behavior to reject more requests, particularly from lower-priority entities. This parameter adjustment maintains service completeness for critical advertisers while preserving overall system performance

Inventive Principle:

Principle #35Parameter changes

4Productivity

If the system rejects requests during overload, then the performance is maintained, but the failure rate increases for rejected requests

Engineering Contradiction:

ImproveperformanceVSAvoidfailure rate

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system applies local quality by ensuring that high-priority advertisers experience minimal to no failures even during overload conditions, while lower-priority entities bear the brunt of rejections. This localized quality assurance maintains high reliability for critical clients while accepting higher failure rates for non-critical ones, thereby preserving overall system performance

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS10783562B2Mitigation of failures in an online advertising network

Publication Date: 2020.09.22 YAHOO AD TECH LLC

AI summary

Access requests to a database are monitored for average time taken to fulfill the requests and whether a queue of unfulfilled requests is building up beyond an acceptable threshold. When the queue has built up beyond the acceptable threshold and/or when the average time taken to fulfill the requests has exceeded a delay threshold value, database access requests may be rejected. In one advantageous aspect, a graceful degradation in performance may be achieved by selectively rejecting access requests of a lower priority and favoring access requests of a higher priority for execution.