Skip to content

An intuitive, easy-to-use python interface for batch resource requesting, access, job submission, and observation. Simplifying the developer's life while enabling access to high-performance compute resources, either in the cloud or on-prem.

License

Notifications You must be signed in to change notification settings

project-codeflare/codeflare-sdk

Repository files navigation

CodeFlare SDK

Python application coverage badge

An intuitive, easy-to-use python interface for batch resource requesting, access, job submission, and observation. Simplifying the developer's life while enabling access to high-performance compute resources, either in the cloud or on-prem.

For guided demos and basics walkthroughs, check out the following links:

  • Guided demo notebooks available here, and copies of the notebooks with expected output also available
  • these demos can be copied into your current working directory when using the codeflare-sdk by using the codeflare_sdk.copy_demo_nbs() function
  • Additionally, we have a video walkthrough of these basic demos from June, 2023

Full documentation can be found here

Installation

Can be installed via pip: pip install codeflare-sdk

Authentication

CodeFlare SDK uses kube-authkit for Kubernetes authentication, supporting multiple authentication methods:

  • Auto-Detection - Automatically detects kubeconfig or in-cluster authentication
  • Token-Based - Authenticate with API server token
  • OIDC - OpenID Connect authentication with device flow or client credentials
  • OpenShift OAuth - Native OpenShift OAuth support
  • Kubeconfig - Traditional kubeconfig file authentication
  • In-Cluster - Service account authentication when running in a pod

Quick Start

from kube_authkit import get_k8s_client, AuthConfig
from codeflare_sdk import set_api_client, Cluster, ClusterConfiguration

# Option 1: Auto-detect authentication (recommended - no explicit auth needed!)
cluster = Cluster(ClusterConfiguration(
    name='my-cluster',
    num_workers=2,
))
cluster.apply()

# Option 2: OIDC authentication
auth_config = AuthConfig(
    method="oidc",
    oidc_issuer="https://your-oidc-provider.com",
    client_id="your-client-id",
    use_device_flow=True
)
api_client = get_k8s_client(config=auth_config)
set_api_client(api_client)  # Register with CodeFlare SDK

# Option 3: OpenShift OAuth with token
auth_config = AuthConfig(
    k8s_api_host="https://api.example.com:6443",
    token="your-token"
)
api_client = get_k8s_client(config=auth_config)
set_api_client(api_client)  # Register with CodeFlare SDK

# Now create your cluster
cluster = Cluster(ClusterConfiguration(
    name='my-cluster',
    num_workers=2,
))
cluster.apply()

Migration from Legacy Authentication

If you're using the deprecated TokenAuthentication or KubeConfigFileAuthentication classes, please see our Migration Guide for detailed instructions on updating to kube-authkit.

Legacy classes (deprecated):

# ⚠️ Deprecated - will be removed in v1.0.0
from codeflare_sdk import TokenAuthentication
auth = TokenAuthentication(token="...", server="...")
auth.login()

New recommended approach:

# ✅ Recommended - Auto-detection (no explicit auth needed!)
from codeflare_sdk import Cluster, ClusterConfiguration
cluster = Cluster(ClusterConfiguration(name="my-cluster"))

# ✅ For OIDC or OpenShift OAuth with token
from kube_authkit import AuthConfig, get_k8s_client
from codeflare_sdk import set_api_client

auth_config = AuthConfig(
    k8s_api_host="https://api.example.com:6443",
    token="your-token"
)
api_client = get_k8s_client(config=auth_config)
set_api_client(api_client)  # Register with CodeFlare SDK

Development

Please see our CONTRIBUTING.md for detailed instructions.

Release Instructions

Automated Releases

It is possible to use the Release Github workflow to do the release. This is generally the process we follow for releases

Manual Releases

The following instructions apply when doing release manually. This may be required in instances where the automation is failing.

About

An intuitive, easy-to-use python interface for batch resource requesting, access, job submission, and observation. Simplifying the developer's life while enabling access to high-performance compute resources, either in the cloud or on-prem.

Resources

License

Contributing

Stars

Watchers

Forks

Packages

No packages published

Contributors 45

Languages