mpinb / cluster-tools

cluster-tools python library from the scalable minds mono-repo

Cluster Tools

This package provides python Executor classes for distributing tasks on a slurm cluster or via multi processing.

Example

import cluster_tools

def square(n):
  return n * n

if __name__ == '__main__':
  strategy = "slurm"  # other valid values are "multiprocessing" and "sequential"
  with cluster_tools.get_executor(strategy) as executor:
    result = list(executor.map(square, [2, 3, 4]))
    assert result == [4, 9, 16]

Configuration

Slurm

The cluster_tools automatically determine the slurm limit for maximum array job size and split up larger job batches into multiple smaller batches. Also, the slurm limit for the maximum number of jobs which are allowed to be submitted by a user at the same time is honored by looking up the number of currently submitted jobs and only submitting new batches if they fit within the limit.

If you would like to configure these limits independently, you can do so by setting the SLURM_MAX_ARRAY_SIZE and SLURM_MAX_SUBMIT_JOBS environment variables.

Kubernetes

Resource configuration

Key	Description	Example
namespace	Kubernetes namespace for the resources to be created. Will be created if not existent.	cluster-tools
node_selector	Which nodes to utilize for the processing. Needs to be a Kubernetes nodeSelector object.	{"kubernetes.io/hostname": "node001"}
image	The docker image for the containerized jobs to run in. The image needs to have the same version of cluster_tools and the code to run installed and in the PYTHONPATH.	scalableminds/voxelytics:latest
mounts	Additional mounts for the containerized jobs. The current working directory and the .cfut directory are automatically mounted.	["/srv", "/data"]
cpu	CPU requirements for this job.	4
memory	Memory requirements for this job. Not required, but highly recommended to avoid congestion. Without resource requirements, all jobs will be run in parallel and RAM will run out soon.	16G
python_executable	The python executable may differ in the docker image from the one in the current environment. For images based of FROM python, it should be python. Defaults to python.	python3.8
umask	umask for the jobs.	0002

Notes

• The jobs are run with the current uid:gid.
• The jobs are removed 7 days after completion (successful or not).
• The logs are stored in the .cfut directory. This is actually redundant, because Kubernetes also stores them.
• Pods are not restarted upon error.
• Requires Kubernetes ≥ 1.23.
• Kubernetes cluster configuration is expected to be the same as for kubectl, i.e. in ~/.kube/config or similar.

Dev Setup

cd dockered-slurm
docker-compose up -d
docker exec -it slurmctld bash
docker exec -it c1 bash

Tests can be executed with cd tests && poetry run pytest -s tests.py after entering the container. Linting can be run with ./lint.sh. Code formatting (black) can be run with ./format.sh.

Credits

Thanks to sampsyo/clusterfutures for providing the slurm core abstraction and giovtorres/slurm-docker-cluster for providing the slurm docker environment which we use for CI based testing.