Sponsored Keynote: OpenSearch at Atlassian Scale
summary
Atlassian runs OpenSearch for 300,000 customers and over 1.5 million monthly active AI users across 13 regions, operating more than 300 clusters with over 2,000 data nodes and a largest cluster holding 23 billion documents. Jira moves from Phosphor full‑text to OpenSearch, requiring migration to eventual consistency. A custom single‑index+routing multi‑tenant workaround adds tenant filtering and visibility gaps, so Atlassian builds a control plane and a Kubernetes‑based platform using Kubernetes + Argo CD + Carpenter, and reports faster iteration and improved observability.
AI-written summary · about
chapters (11)
- 0:00Atlassian’s OpenSearch journey
- 1:04OpenSearch use cases
- 2:30Rovo AI capabilities
- 3:36Scale challenges with OpenSearch
- 5:45Challenges into opportunities
- 7:30Centralized infrastructure shift
- 9:16Innovation vs operational demands
- 10:41Kubernetes-based OpenSearch infrastructure
- 12:47Self-managed OpenSearch benefits
- 13:48Stateful services on Kubernetes
- 15:33Community and future opportunities
related talks
43mMulti-tenancy for all workloads
35mHighly Available Search at Shopify
43mDeep Dive Into OpenSearch Pull-Based Ingestion at Uber
14mJourney of Building a Scalable, Domain-Agnostic Order Management System
19mYou Know, for Search. Querying 24 Billion Records in 900ms
31mSharding 2.0: Unlocking Infinite Search Scalability in OpenSearch
24mExtending OpenSearch: A Practical Guide To Building Custom Plugins
40mNon-Obvious Things to Know When Coding in OpenSearch