Server Overload Detection via Request Buffer Evaluation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Data service servers face overload issues due to sudden spikes in client requests, leading to server restarts even when the service is functional, which further strains other servers and can cause service failures.

Innovation Solution

Implementing an overload detection mechanism that evaluates the request buffer condition to determine if a client request corresponds to an anticipated overload state, returning a 'retry-after-delay' response to clients if detected, allowing the server to process outstanding requests without immediate restarts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If the server processes all incoming client requests during high load periods, then the service maintains high responsiveness to clients, but the server becomes overloaded and fails heartbeat detection causing unnecessary restarts

Engineering Contradiction:
Improverequest processing capacityVSAvoidheartbeat detection reliability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The server performs preliminary evaluation of the request buffer condition before accepting new requests. When the buffer condition indicates anticipated overload, the server proactively rejects requests with a 509 response code before the overload actually occurs, preventing the server from entering a failed state while maintaining reliability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The server continuously monitors its request buffer condition and uses this feedback to dynamically adjust its request acceptance behavior. When the buffer condition exceeds thresholds, the server adjusts by rejecting requests, creating a feedback loop that maintains operational reliability under varying load conditions.

Inventive Principle:
Principle #23Feedback

2Reliability

If the server rejects incoming requests during high load to avoid overload, then the server maintains operational stability and passes heartbeat detection, but client service experience deteriorates

Engineering Contradiction:
Improveserver operational stabilityVSAvoidclient service accessibility
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The server dynamically adjusts its request handling behavior based on real-time buffer conditions rather than using a static approach. The server transitions between accepting and rejecting requests based on the current load state, optimizing both stability and client service experience through adaptive behavior.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The server changes the parameter of request acceptance based on the buffer condition evaluation. When the buffer condition indicates safety, the server accepts requests; when it indicates anticipated overload, the server changes the parameter to reject requests with 509 responses, optimizing the balance between stability and service accessibility.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If the server adds more server instances to handle increased load, then the service can handle more requests, but the infrastructure complexity and cost increase

Engineering Contradiction:
Improverequest handling capacityVSAvoidserver infrastructure complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The solution segments the request handling capacity into two categories: requests that can be buffered and processed later, and requests that must be rejected immediately due to anticipated overload. This segmentation allows the existing server infrastructure to handle variable loads without requiring proportional increases in server instances.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3808054B1Data service overload detection and mitigation
Publication Date: 2024.03.06 HOME BOX OFFICE INC
  • EP3808054B1 patent drawingFigure 1
  • EP3808054B1 patent drawingFigure 2
  • EP3808054B1 patent drawingFigure 3

AI summary

The described technology is directed towards presenting a server overload condition that can result from a spike in client requests. Upon receiving an incoming client request for data at a server, a request buffer condition of a client request buffer (e.g., a counter) is evaluated to determine whether the client request corresponds to an anticipated overload state with respect to a request-handling capability of the server, e.g., whether the client request buffer is full. If so, an overload response (e.g., a retry response) is returned to the client indicating so that the client retries a corresponding client request after a time delay. If not in the anticipated overload state, the request is further processed, e.g., to provide a regular response containing the requested data, such as by adding a request information entry associated with the incoming client request to the client request buffer and updating the request buffer condition.