Packet Fetch Scheduler Bandwidth Utilization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In data communication switches, bandwidth is wasted due to packet drops at egress ports, leading to inefficiencies in bandwidth utilization and increased latency, particularly in incast scenarios where output port bandwidth is overallocated, causing bottlenecks and congestion.
Innovation Solution
Implementing a fetch scheduler that manages packet segment fetch rates and uses overspeed to fetch headers before payloads, allowing concurrent fetching of multiple packets in an interleaved manner, and dynamically adjusting fetch rates based on drop rates and congestion levels, with options for store-and-forward or virtual cut-through modes to optimize bandwidth utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If packets are fetched from switch fabric at egress port, then bandwidth utilization is improved, but bandwidth is wasted when packets are dropped
Solution Approach 1:
The patent applies preliminary action by checking packet drop conditions at ingress ports before packets are fetched from the switch fabric. The ingress packet processor determines whether packets should be dropped based on egress port status, and this information is used to prevent fetching packets that would subsequently be dropped, thereby avoiding bandwidth waste while maintaining efficient bandwidth utilization for packets that will actually be transmitted
Solution Approach 2:
The patent extracts the packet drop determination function from the egress port and relocates it to the ingress port. By extracting this decision-making capability to the ingress side, the system can identify and eliminate packets destined for dropped transmissions before they consume switch fabric bandwidth, separating the useful packet transmission from the harmful bandwidth consumption
2Productivity
If packets are fetched from switch fabric, then egress bandwidth is utilized, but latency increases due to gaps between packets
Solution Approach 1:
The patent ensures continuity of useful action by preventing gaps in packet transmission caused by dropped packets. By determining drop conditions at ingress and selectively preventing fetch requests for packets that would be dropped, the system maintains continuous utilization of egress bandwidth without interruption, eliminating latency gaps that would otherwise occur when packets are fetched and then dropped at egress
3Productivity
If E-ACL drop conditions are checked at ingress, then packet fetch bandwidth is optimized, but device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the packet processing function into distinct components: the ingress packet processor that handles drop condition determination, and the egress pipeline that handles packet transmission. This segmentation allows the drop determination logic to be implemented as a separate modular function at ingress, reducing the complexity burden on any single component while optimizing overall packet fetch bandwidth efficiency
Data Source
AI summary
Examples describe a manner of scheduling packet segment fetches at a rate that is based on one or more of: a packet drop indication, packet drop rate, incast level, operation of queues in SAF or VCT mode, or fabric congestion level. Headers of packets can be fetched faster than payload or body portions of packets and processed prior to queueing of all body portions. In the event a header is identified as droppable, fetching of the associated body portions can be halted and any body portion that is queued can be discarded. Fetch overspeed can be applied for packet headers or body portions associated with packet headers that are approved for egress.