Large Instances Meeting Notes 2025-09-16
Transcript: Kubernetes & Large/Multi Instances DevOps collaboration - 2025/09/16 13:02 EDT - Transcript
Recording: Kubernetes & Large/Multi Instances DevOps collaboration - 2025/09/16 13:02 EDT - Recording
AI Summary:
Updates
Braden MacDonald
No major updates.
Team continues exploring the PHD stack (Picasso, Harmony, Drydock) migration.
Began benchmarking Typesense as an alternative to Meilisearch.
Felipe Montoya
Collaborating with the Picasso team on making Tutor data handling more flexible (e.g., config file locations).
Working on AI extensibility plugin development; expects to push this further in Oct–Nov and onboard contributors.
Maksim Sokolskiy
No major updates this cycle, mostly in read-only mode.
Presented Felipe’s AI repo to someone who may begin exploring it.
Moisés González
At an internal hackathon, tested the MariaDB Operator for Kubernetes:
Deploys robust MySQL-compatible DBs within a namespace.
Supports high availability (Galera, primary/secondary), multiple backup types, and integrates with monitoring.
Successfully prototyped deployment with Open edX.
Broader Discussion
Cloudflare:
Felipe raised its relevance due to AI scraping controls; attendees had only limited free-tier experience, no large-scale production usage reported.
Nginx & ACME:
Recent improvements in native ACME certificate handling noted.
Discussion on past migration from Caddy → Nginx → Ingress in Tutor.
Some concerns about future compatibility but overall stable.
Kubernetes Gateway API:
Braden and Moisés discussed how Gateway aims to replace Ingress for better flexibility and security.
Next Steps
Felipe to advance AI plugin work and revisit in Oct/Nov.
Explore whether the MariaDB Operator could improve Open edX hosting infrastructure.
Keep monitoring Kubernetes and ingress/gateway developments for future implications.
Meeting length: ~15 minutes.
Next meeting: In two weeks.