Adaptive Storm Control via Credit-Based Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network performance is degraded by traffic storms, where excessive packet flooding is not effectively managed, leading to unnecessary traffic drops during legitimate network changes or peak usage periods.
Innovation Solution
An adaptive storm control system that employs a credit-based mechanism to allow temporary rate increases during predefined conditions, such as peak usage times, while limiting traffic to a steady state rate when conditions are not met, using a storm controller to manage packet forwarding based on thresholds and available credits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional storm control drops all traffic exceeding the threshold, then network stability is improved, but legitimate network convergence traffic is unnecessarily blocked
Solution Approach 1:
The storm control system dynamically adjusts its behavior based on traffic conditions and predefined criteria. Instead of using a fixed threshold policy, the system evaluates whether traffic bursts represent legitimate network events (such as convergence traffic) or actual storms, and adapts its forwarding decisions accordingly. This allows the system to maintain stability for genuine storms while permitting necessary network operations.
Solution Approach 2:
The system changes the operational parameters of storm control by introducing a credit-based mechanism that modifies the effective threshold behavior. When credits are available, the system effectively raises the threshold to allow burst traffic; when credits are depleted, it returns to strict threshold enforcement. This parameter adjustment resolves the contradiction by making the control mechanism flexible rather than rigid.
2Object-affected harmful factors
If storm control enforces strict threshold limits, then traffic storms are effectively suppressed, but network performance during peak usage deteriorates
Solution Approach 1:
The system performs preliminary actions by accumulating credits during periods of low traffic demand. These pre-accumulated credits serve as a reserve that can be spent during peak usage periods or legitimate network events, allowing the system to temporarily exceed threshold limits without immediate penalty. This preliminary preparation enables the system to distinguish between malicious storms and legitimate high-demand traffic.
Solution Approach 2:
The credit-based mechanism implements a feedback system where the storm controller continuously monitors traffic patterns and adjusts forwarding decisions based on available credits. When traffic patterns indicate legitimate usage (such as network convergence or peak demand), the system feedback allows threshold exceedance if credits permit. This feedback loop prevents false suppression of beneficial traffic while maintaining storm control effectiveness.
3Loss of energy
If traffic is dropped during storm conditions, then network resource consumption is reduced, but legitimate network events are blocked
Solution Approach 1:
The system segments the traffic handling approach by separating credit evaluation from threshold evaluation. Instead of a single monolithic decision process, the system first checks credit availability, then applies threshold rules conditional on credit status. This segmentation allows differentiated treatment of traffic: strict suppression when credits are depleted, and flexible permitting when credits are available, thereby handling both resource conservation and legitimate events appropriately.
Solution Approach 2:
The credit system acts as an intermediary mechanism between the storm control threshold and the actual packet forwarding decision. Rather than directly blocking or forwarding packets based solely on threshold exceedance, the credit intermediary evaluates whether the threshold violation represents a legitimate event. This intermediary layer adds adaptability by mediating between resource conservation goals and network functionality requirements.
Data Source
AI summary
In an example embodiment, there is disclosed herein an apparatus comprising an ingress interface, an egress interface, and a storm controller coupled to the ingress interface and the egress interface. The storm controller is operable to determine whether to forward packets for a traffic flow received at the ingress interface to the egress interface based on a rate over a time period. The storm controller forwards packets for the traffic flow while the rate exceeds a first threshold and is less than a second threshold while a predefined condition exits. The storm controller limits traffic for the traffic flow to the first threshold while the rate exceeds the first threshold and the predefined condition does not exist.