Adaptive Network Packet Egress Scheduling for Bandwidth Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network compute devices face challenges in efficiently managing network traffic due to varying port speeds and packet sizes, leading to higher power consumption and design complexities, as they are often provisioned for minimum packet sizes, resulting in suboptimal bandwidth utilization and increased clock frequencies.
Innovation Solution
The implementation of an adaptive network packet egress scheduling system that dynamically adjusts the scheduling sequence to favor more utilized ports, shifting bandwidth from less utilized to more utilized ports, using a minimum bandwidth allocation table and eligible mask to determine which ports to service, thereby optimizing the number of ports that can be serviced with available fabric bandwidth.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the scheduler is fully provisioned for minimum sized packets on all ports, then all ports can maintain their line rate and avoid overrun/underrun, but the clock frequency and power consumption increase
Solution Approach 1:
The patent implements dynamic scheduling where the scheduler adapts its behavior based on actual traffic conditions. Instead of statically provisioning for minimum packet sizes on all ports, the scheduler dynamically adjusts service rates and port servicing frequency based on real-time queue depths and traffic patterns, allowing the system to maintain reliability while reducing power consumption during low-traffic periods
Solution Approach 2:
The system changes operational parameters dynamically - specifically adjusting the service rate, port servicing frequency, and chunk size based on actual traffic conditions. The scheduler modifies these parameters in response to varying packet sizes and traffic patterns, allowing the fabric to operate at optimal clock frequencies rather than being locked to maximum provisioning levels
2Reliability
If the scheduler services ports at higher frequencies to handle minimum sized packets, then packet loss is avoided, but bandwidth utilization decreases and design complexity increases
Solution Approach 1:
The scheduler dynamically adjusts the service rate for each port based on actual traffic conditions rather than using fixed high-frequency servicing. The system monitors queue depths and traffic patterns, adjusting service rates to match actual demand, thereby preventing packet loss while maximizing bandwidth utilization by avoiding unnecessary high-frequency servicing during low-traffic periods
Solution Approach 2:
Instead of always servicing all ports at maximum frequency, the system applies partial servicing - only servicing ports that have packets in their queues and adjusting the frequency based on actual needs. This avoids the excessive action of continuously servicing empty queues at high frequencies, thereby improving bandwidth utilization while maintaining packet loss prevention
Data Source
AI summary
Technologies for adaptive network packet egress scheduling include a switch configured to configure an eligibility table for a plurality of ports of the switch, wherein the eligibility table includes a plurality of rounds. The switch is further configured to retrieve an eligible mask corresponding to a round of a plurality of rounds of the eligibility table presently being scheduled and determine a ready mask that indicates a ready status of each port. The switch is further configured to determine, for each port, whether the eligible status and the ready status indicate that port is both eligible and ready, and schedule, in response to a determination that at least one port has been determined to be both eligible and ready, each of the at least one port that has been determined to be both eligible and ready. Additional embodiments are described herein.