Architecture 15 - Galaxy Production Deployment

Contributors

John Chilton

Questions

• How is Galaxy deployed in production?

• What is the difference between development and production setups?

• How does usegalaxy.org work?

Objectives

• Understand production deployment architecture

• Learn about PostgreSQL, nginx, and uWSGI

• Understand multi-process and multi-host setups

• Learn about usegalaxy.org infrastructure

layout: introduction_slides topic_name: Galaxy Architecture

Architecture 15 - Galaxy Production Deployment

The architecture of Galaxy’s production deployment.

layout: true left-aligned class: left, middle — layout: true class: center, middle

class: centered

.pull-left[

Default

SQLite

gunicorn all-in-one

Single process

Single host

Local jobs

]

.pull-right[

Production

PostgreSQL

gunicorn for web process + webless workers + nginx proxy

Multiple processes

Multiple hosts

Jobs across many clusters

https://usegalaxy.org/production ]

PostgreSQL

• Database server can scale way beyond default sqlite
• Supports concurrent connections from multiple Galaxy processes
• Better performance for production workloads
• https://www.postgresql.org/
• Configuration: github.com/galaxyproject/usegalaxy-playbook → roles/galaxyprojectdotorg.postgresql

nginx (or Apache)

• Optimized servers for serving static content
• Reverse proxy to Galaxy application servers
• Load balancing across multiple Galaxy processes
• https://www.nginx.com/resources/wiki/
• https://docs.galaxyproject.org/en/master/admin/nginx.html#proxying-galaxy-with-nginx
• Configuration: github.com/galaxyproject/usegalaxy-playbook -> templates/nginx/usegalaxy.j2

Webless

• Galaxy typically runs in Gunicorn - a production-grade ASGI server
• This is a great tool for both development and production but in production typically job running and workflow scheduling should happen outside a webserver
• https://docs.galaxyproject.org/en/master/admin/scaling.html#gunicorn-for-web-serving-and-webless-galaxy-applications-as-job-handlers
• https://training.galaxyproject.org/training-material/topics/admin/tutorials/ansible-galaxy/tutorial.html

Multi-processes

Threads in Python are limited by the GIL.

Running multiple processes of Galaxy and separate processes for web handling and job processing works around this.

This used to be an important detail - but gravity + gunicorn make things a lot easier.

class: center

Cluster Support

Galaxy can submit jobs to various cluster managers (Slurm, PBS, SGE, etc.)

https://docs.galaxyproject.org/en/master/admin/cluster.html

https://training.galaxyproject.org/training-material/topics/admin/tutorials/connect-to-compute-cluster/tutorial.html

class: center

usegalaxy.org Web Architecture

class: center

Complete usegalaxy.org Infrastructure

Multiple web servers, job handlers, and compute clusters working together

Key Production Considerations

• Database: Use PostgreSQL for production
• Web Server: Use nginx or Apache as reverse proxy
• Processes: Run multiple Galaxy processes
• Job Handling: Separate job handlers from web workers
• Storage: Use scalable object storage solutions
• Monitoring: Implement logging and monitoring
• Backups: Regular database and file backups

Production Deployment Resources

• Admin Training: https://training.galaxyproject.org/topics/admin/
• Galaxy Admin Docs: https://docs.galaxyproject.org/en/master/admin/
• Ansible Playbooks: https://github.com/galaxyproject/usegalaxy-playbook
• Community Support: https://help.galaxyproject.org/

Key Points

• Production uses PostgreSQL instead of SQLite
• nginx/Apache serve static content, uWSGI runs Galaxy
• Multiple processes and hosts for scalability
• usegalaxy.org runs across multiple servers and clusters

Thank you!