kblaunch 0.2.20

A CLI tool for launching Kubernetes job fast in EIDF

kblaunch

A CLI tool for launching Kubernetes jobs with environment variable and secret management.

Installation

pip install kblaunch

Or using uv:

uv add kblaunch

You can even use uvx to use the cli without installing it:

uvx kblaunch --help

Usage

Setup

Run the setup command to configure the tool (email and slack webhook):

kblaunch setup

This will go through the following steps:

1. Set the user (optional): This is used to identify the user and required by the cluster. The default is set to $USER.
2. Set the email (required): This is used to identify the user and required by the cluster.
3. Set up Slack notifications (optional): This will send a test message to the webhook, and setup the webhook in the config. When your job starts you will receive a message at the webhook
4. Set up a PVC (optional): This will create a PVC for the user to use in their jobs
5. Set the default PVC to use (optional): Note only one pod can use the PVC at a time

Basic Usage

Launch a simple job:

kblaunch launch
    --job-name myjob \
    --command "python script.py"

With Environment Variables

1. From local environment:

   export PATH=...
   export OPENAI_API_KEY=...
   # pass the environment variables to the job
   kblaunch launch \
       --job-name myjob \
       --command "python script.py" \
       --local-env-vars PATH \
       --local-env-vars OPENAI_API_KEY
   

2. From Kubernetes secrets:

   kblaunch launch \
       --job-name myjob \
       --command "python script.py" \
       --secrets-env-vars mysecret1 \
       --secrets-env-vars mysecret2
   

3. From .env file (default behavior):

   kblaunch launch \
       --job-name myjob \
       --command "python script.py" \
       --load-dotenv
   

   If a .env exists in the current directory, it will be loaded and passed as environment variables to the job.

GPU Jobs

Specify GPU requirements:

kblaunch launch \
    --job-name gpu-job \
    --command "python train.py" \
    --gpu-limit 2 \
    --gpu-product "NVIDIA-A100-SXM4-80GB"

Interactive Mode

Launch an interactive job:

kblaunch launch \
    --job-name interactive \
    --interactive

Launch Options

Launch command options:

• --email: User email (overrides config)
• --job-name: Name of the Kubernetes job [required]
• --docker-image: Docker image (default: "nvcr.io/nvidia/cuda:12.0.0-devel-ubuntu22.04")
• --namespace: Kubernetes namespace (default: "informatics")
• --queue-name: Kueue queue name (default: "informatics-user-queue")
• --interactive: Run in interactive mode (default: False)
• --command: Command to run in the container [required if not interactive]
• --cpu-request: CPU request (default: "1")
• --ram-request: RAM request (default: "8Gi")
• --gpu-limit: GPU limit (default: 1)
• --gpu-product: GPU product type (default: "NVIDIA-A100-SXM4-40GB")
  • Available options:
    • NVIDIA-A100-SXM4-80GB
    • NVIDIA-A100-SXM4-40GB
    • NVIDIA-A100-SXM4-40GB-MIG-3g.20gb
    • NVIDIA-A100-SXM4-40GB-MIG-1g.5gb
    • NVIDIA-H100-80GB-HBM3
• --secrets-env-vars: List of secret environment variables (default: [])
• --local-env-vars: List of local environment variables (default: [])
• --load-dotenv: Load environment variables from .env file (default: True)
• --nfs-server: NFS server address
• --pvc-name: Persistent Volume Claim name
• --dry-run: Print job YAML without creating it (default: False)
• --priority: Priority class name (default: "default")
  • Available options: default, batch, short
• --vscode: Install VS Code CLI in container (default: False)
• --tunnel: Start VS Code SSH tunnel on startup (requires SLACK_WEBHOOK and --vscode)
• --startup-script: Path to startup script to run in container

Monitor command options:

• --namespace: Kubernetes namespace (default: "informatics")

Monitoring Commands

The kblaunch monitor command provides several subcommands to monitor cluster resources:

Displays aggreate GPU statistics for the cluster:

kblaunch monitor gpus

Displays queued jobs (jobs which are waiting for GPUs):

kblaunch monitor queue

Displays per-user statistics:

kblaunch monitor users

Displays per-job statistics:

kblaunch monitor jobs

Note that users and jobs commands will run nvidia-smi on pods to obtain GPU usage is not recommended for frequent use.

Features

• Kubernetes job management
• Environment variable handling from multiple sources
• Kubernetes secrets integration
• GPU job support
• Interactive mode
• Automatic job cleanup
• Slack notifications (when configured)
• Persistent Volume Claim (PVC) management
• VS Code integration (with Code tunnelling support)
• Monitoring commands