PLUTUS : Serverless monitoring and alerting of costs incurred on AWS
summary
A near‑real‑time cost monitoring pipeline tracks AWS Cost and Usage Reports into S3, imports CSVs into Redshift, runs resource‑level queries and sends daily/weekly alerts to Slack via Lambda, targeting volatile items like Autoscaling Groups, untracked Hadoop jobs, idle on‑demand EC2 and high S3 egress. The claim is concrete: Plutus programmatically detects threshold breaches, stabilizes monthly bills and reports roughly $2M/year saved after deployment.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:02AWS infrastructure overview
- 2:06Cost control challenges
- 3:51Spot instance savings
- 5:33Cost spike impacts
- 6:36Need for custom monitoring
- 8:01PLUTUS and Slack alerts
- 9:49Billing to Redshift pipeline
- 11:58Lambda automation
- 12:41Results and open source
- 13:23Q&A
related talks
36mFinOps & Observability: technical implementation
31mThe AWS Billing Machine and Optimizing Cloud Costs
31mHow (Not) to Spend $15000 per Month on AWS Athena
27mMastering AWS Cost Optimisation
20mCloud Cost Management Updates
42mServerless Tales from the Trenches
19mUltra-Budget AWS: Running Scalable Apps for Pennies
46mWhen the CFO Comes Knocking: Our Journey to a Cost Efficient Data Platform