Krake — Workload scheduling across cloud boundaries

Krake is an open-source workload manager for distributed infrastructures. It orchestrates the efficient distribution of virtualized and containerized workloads across multi-cloud, private cloud, and on-premises environments, and selects target systems based on freely definable environmental, technical, or economic metrics.

Where to find it

• Code gitlab.com/alasca.cloud/krake/krake
• Docs rak-n-rok.readthedocs.io/projects/krake/en/latest
• Website krake.cloud
• Issues gitlab.com/groups/alasca.cloud/krake/-/work_items

Maintained by

• Patrick Thiem Project representative · Cloud & Heat

Written in

PythonJinja

Licence

Apache License 2.0

Context

• Multi-Cloud
• Kubernetes
• OpenStack
• Scheduling

Presentations

• ALASCA Tech Talk #7: Unleashing the Krake(n) for Optimal Kubernetes Workload Management

Project Statistics

The data is gathered automatically and periodically. Beware that there might be issues on how they have been collected. The data was lastly gathered on 2026-10-21. Every contribution, organisation and approval counted below falls within the last 12 months.

ActivityOpen/closed issuesOpen/closed merge requests

Metric	Count
Commits (365d)	0
Total merged merge requests	346
Open merge requests	8
Open Issues	86
Forks	4
Stars	27
Releases (365d)	0

Protocols

• 2025-08-22 Krake Introduction with Patrick Thiem