Autonomous Multi-Cloud serverless deployment
summary
Melodic is an open‑source (MPL 2.0) multi‑cloud deployment and optimization platform that models apps with the Kamal language, provisions VMs/containers/serverless across providers, and evaluates candidates with a configurable utility function based on collected metrics (Grafana used for dashboards). Demo deploys a Spark “genome” job on AWS, auto‑scales workers from 1 to 2 when estimated time exceeds the target, and completes successfully. Future features include polymorphic architecture and proactive adaptation forecasting workloads; one use case claims ~$107k savings over three years.
AI-written summary · about
chapters (13)
- 0:00Introduction to MELODIC
- 1:33Why use MELODIC platform
- 4:45How MELODIC works
- 6:30Use case Investments
- 8:37Live demo setup
- 9:00Deploying Spark application
- 12:37MELODIC configuration
- 14:22Deployment process overview
- 18:03Reconfiguration for optimization
- 20:32Monitoring with Grafana
- 24:05Reconfiguration in action
- 25:31Demo conclusion
- 26:16Resources and community
related talks
29mProactive and Polymorphic Adaptation of Multi-cloud Deployments
49mProactive optimization for Cloud
13mSam Kroonenburg, A Cloud Guru | ServerlessConf 2018
24mWhat it Takes to Build & Operate a Global Serverless Provider
24mBeyond Tool Sprawl: An Event-Driven Approach to Dynamic Multi-Cloud Operations
15mJourney to Hybrid Multi-Cloud
27mFrom Supercomputing to Serving: A Case Study Delivering Cloud Native Foundation Models
17mCloud Computing Made Better: An AWS View on Serverless Computing