CPU Feature Discovery for Kubernetes

Overview
Getting Started
- System Requirements
- Usage
Building from source
Targeting Nodes with Specific Features
License

Overview

This software enables Intel Architecture (IA) feature discovery for Kubernetes. It detects CPU features available on each node in a Kubernetes cluster, such as Intel Resource Director Technology and advertises those features using node labels.

Intel Resource Director Technology (RDT) Features

Feature name	Description
RDTMON	Intel Cache Monitoring Technology (CMT) and Intel Memory Bandwidth Monitoring (MBM)
RDTL3CA	Intel L3 Cache Allocation Technology
RDTL2CA	Intel L2 Cache Allocation Technology

Other Features (Partial List)

Feature name	Description
ADX	Multi-Precision Add-Carry Instruction Extensions (ADX)
AESNI	Advanced Encryption Standard (AES) New Instructions (AES-NI)
AVX	Advanced Vector Extensions (AVX)
AVX2	Advanced Vector Extensions 2 (AVX2)
BMI1	Bit Manipulation Instruction Set 1 (BMI)
BMI2	Bit Manipulation Instruction Set 2 (BMI2)
SSE4.1	Streaming SIMD Extensions 4.1 (SSE4.1)
SSE4.2	Streaming SIMD Extensions 4.2 (SSE4.2)
SGX	Software Guard Extensions (SGX)

The published node labels encode a few pieces of information:

A "namespace" to denote vendor-specific information (node.alpha.intel.com).
The version of this discovery code (e.g. v0.1.0) that wrote the label.
The relevant hardware component each label describes (e.g. cpu).
The name of the discovered feature as it appears in the underlying source, mostly cpuid (e.g. AESNI).

Note: only features that are available on a given node are labeled, so the only label value published is the string "true". This feature discovery code will not add a label with the value "false" for features that are not present.

"node.alpha.intel.com/v0.1.0/cpu/<feature-name>": "true"

System Requirements

At a minimum, you will need:

Linux (x86_64)
[kubectl] kubectl-setup (properly set up and configured to work with your Kubernetes cluster)
[GCC] gcc-down (only required to build software to detect Intel RDT feature set)
[Docker] docker-down (only required to build and push docker images)

Usage

Feature discovery is done as a one-shot job. There is an example script in this repo that demonstrates how to deploy the job to unlabeled nodes.

./label-nodes.sh

The discovery script will launch a job on each each unlabeled node in the cluster. When the job runs, it contacts the Kubernetes API server to add labels to the node to advertise hardware features (initially, from cpuid and RDT).

Building from source

Download the source code.

git clone https://github.com/intelsdi-x/dbi-iafeature-discovery

The build steps described below are optional. The default docker image in Dockerhub at intelsdi/nodelabels can be used to decorate the Kubernetes node with CPU features. Skip to usage instructions if you do not need to build your own docker image.

Build the Intel RDT Detection Software Using make (Optional):

cd <project-root>/rdt-discovery
make

Build the Docker image (optional):

cd <project-root>
docker build -t <user-name>/<image-name> .

Push the Docker Image (optional)

docker push

Change the job spec to use your new image (optional):

To use your published image from the step above instead of the intelsdi/nodelabels image, edit line 40 in the file [dbi-iafeature-discovery-job.json.template](dbi-iafeature-discoverz-job.json.template) to the new location (<user>/<image-name>).

Targeting Nodes with Specific Features

Nodes with specific features can be targeted using the nodeSelector field. The following example shows how to target the Intel RDT L3 cache allocation (RDTL3CA) feature.

{
    "apiVersion": "v1",
    "kind": "Pod",
    "metadata": {
        "labels": {
            "env": "test"
        },
        "name": "golang-test"
    },
    "spec": {
        "containers": [
            {
                "image": "golang",
                "name": "go1",
            }
        ],
        "nodeSelector": {
                ""node.alpha.intel.com/v0.1.0-cpu-RDTL3CA": "true"
        }
    }
}

For more details on targeting nodes, see node selection.

License

This is open source software released under the Apache 2.0 License.

5.4 KiB Raw Blame History