Peak I/O performance for different methods of INC Input Data transfer

[robcraig-LF]

Classification Tag

Technical

Impact Level

High (functionality or security issues)

Affected Section(s)

Current document misreports the drop percentage as 5%., when throughput actually drops by 2% (or 1%) when an additional minimum size packet is needed for each 4096 (or 8192) Byte MTU using standard header and IPG. This can be improved using smaller min-size frame or using cumulative ACK (which requires additional Rx buffer).

Original Text

Throughput drops by 5% when an additional minimum size packet is needed for each 4096 (or 8192) Byte MTU using standard header and IPG. This can be improved using smaller min-size frame or using cumulative ACK (which requires additional Rx buffer)

Proposed Resolution

Throughput drops by 2% (or 1%) when an additional minimum size packet is needed for each 4096 (or 8192) Byte MTU using standard header and IPG. This can be improved using smaller min-size frame or using cumulative ACK (which requires additional Rx buffer)

Justification & Additional Context

This is the correct percentage as tested.

Maintainer Checklist

• Classification and impact reviewed
• Section reference verified
• Assigned to appropriate team #68
• Version tag applied