NoOps for OpenSearch: The Future of Self-Healing Search Infras...
summary
Shows a working EKS proof of concept that wires OpenSearch via EKS + Helm + Terraform with the required EBS CSI driver, deploys a single-replica OpenSearch (1GB RAM, ~500m CPU, 1GB PVC) and demonstrates self-healing: deleted pods are immediately recreated, plus HPA triggers at 50% CPU and scales up to 3 replicas under load generated by a curl loop. Covers Kubernetes cluster autoscaler, automated upgrades, observability gaps and cost/complexity tradeoffs; demo succeeds but highlights upgrade and monitoring challenges.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:43Why NoOps is needed
- 3:28Manual operations impact
- 4:10Infrastructure complexity
- 4:52Resilience and agility demand
- 5:32Faster time to market
- 6:12Search infrastructure evolution
- 7:58What is OpenSearch
- 9:02OpenSearch power and pain
- 9:44NoOps vision for OpenSearch
- 10:48Core NoOps enablers
- 11:51Self-healing explained
- 13:14Autoscaling with Kubernetes
- 14:15Cluster management challenges
- 14:58Demo setup for self-healing
- 17:45Auto-healing in action
- 18:28Load testing autoscaling
- 19:53NoOps benefits
- 20:14Q&A
related talks
42mOpenSearch in Kubernetes: Helm Charts and Operator
23mCreating the OpenSearch Kubernetes Operator
52mOperating OpenSearch at Scale: Lessons From Managing 10,000+ Clusters Across...
39mOperating OpenSearch at Scale With Kubernetes Operators: Lessons Learned and Community Insights
36mAutoscaling Search and Analytics in Kubernetes: The OpenSearch Operator Advantage
38mMigrating from Self-Managed ElasticSearch to OpenSearch
30mFrom Chaos To Control: GitOps Disaster Recovery With the OpenSearch Operator
14mKeynote: OpenSearch on Kubernetes at Atlassian