Refer to the Onload User Guide (UG1586) to compare all standard tunables. Many configurations, such as kernel boot parameters, NUMA node selection, and Onload profiles, are specified in slightly different ways.
This is likely caused by a variable other than Onload Operator for Kubernetes as:
- Onload Operator does not introduce any additional layers into the Onload bypass stack; it is only orchestration.
- The version of Onload compiled in the onload-user images and the kernel modules compiled in-cluster or out of cluster from the onload-source image match the standard Linux packages and are not customized variants.
- The kernel module files can be encapsulated in container images but are loaded directly and natively into the kernel.
Contact your AMD sales representative and support team to discuss optimization of your workloads. See also the Release Notes from the Xilinx-CNS/kubernetes-onload repository.