OpenSearchCon India 2026

We want to show that a fully open-source log management stack (OpenSearch + Data Prepper + OpenSearch Dashboards) can run at ~400TB-per-day scale with few extensions. We hope to get feedback from others running ODD at scale, exchange patterns for plugin development and operations, and encourage more teams to adopt and contribute to the ODD ecosystem.

It details a massive ODD deployment utilizing ~1,500 OpenSearch nodes to handle 6 GB/s data volume. Key takeaways are Extensibility (Pluggable data prepper), Cost Efficiency (S3 remote snapshots as cold storage), and Open Source Commitment (using official OpenSearch ecosystem). Attendees will grasp ODD scaling and extension potential.

It showcases how the ODD stack manages high-throughput pipelines and multi-datasource UIs, reducing reliance on proprietary vendors. It presents extension patterns using Data Prepper and OSD plugins for custom observability (RUM, synthetics, traces, profiles, alerts). This promotes open source observability through a reusable reference architecture and operational patterns, encouraging OpenSearch ecosystem adoption and contribution.