Feature Set N-ANL11

Intel® PAC
Napatech SmartNIC
Content Type
Feature Description

This section describes principles of operation of the correlation key functionality.

Data selection for correlation key calculation

To prevent for example dynamic header information from affecting the correlation key calculation, the subset of the frame included in the calculation of the correlation key is configured by dynamic start and end positions, combined with masking functionality.

Data selection for correlation key calculation

IP Fragments

To ensure that the correlation key is calculated for the same range if a frame is being fragmented during transport, a maximum length of the range used for calculation of the correlation key can be specified. This allows for example correlation of an unfragmented IP packet with the first fragment of that packet at a downstream node in a network.

Correlation of non-fragmented packet with first fragment

Calculation of correlation key

A 64-bit CRC is calculated for the selected subset of the frame, and the result is used as the correlation key.

The 64-bit CRC polynomial used is: x64 + x4 + x3 + x + 1.

Packet descriptors

The calculated correlation key is available to the host application when one of the following packet descriptors is used:
  • DYN4: the complete 64-bit correlation key
  • EXT9: the lower 24 bits of the correlation key