Operating OpenSearch at Scale: Lessons From Managing 10,000+ Clusters Across...

Operating OpenSearch at Scale: Lessons From Managing 10,000+ Clusters Across Hyperscalers - Hariharan Gandhi, SAP SE Running OpenSearch as a managed service at massive scale is no small feat. In this session, we’ll explore the architectural and operational strategies behind hosting tens of thousands of OpenSearch clusters productively across heterogeneous environments. We’ll share how we built a flexible, K8s-based platform capable of supporting a wide range of cluster sizes. Using custom operators, we automate critical lifecycle tasks such as rolling updates, bootstrapping, scaling, and day-2 operations—including ISM, retention policies, automated instance maintenance and deliver curated content bundles to users, keeping environments consistent and up-to-date. Beyond the basics, we’ll explore some of our ongoing investigations and challenges—such as extra-large clusters, cross-cluster search, controlled egress, and compliance—which could provide food for thought and spark valuable discussion after the session. Attendees will walk away with: Strategies & Practical insights for managing OpenSearch at scale using Kubernetes operators Whether you’re running OpenSearch internally or offering it as a platform service, this talk offers actionable insights drawn from real-world experience at scale.