Rajith Attapattu
Open Telemetry advocate - CTO @ Randoli
Toronto, Canada
Actions
Rajith is an advocate for OpenTelemetry and open source as a member of the Apache Software Foundation. He spends his time experimenting, contributing, and exploring new ways to improve observability. As CTO of Randoli, he’s building an OTel native observability platform to help teams gain deeper insights. Previously, he spent 11 years at Red Hat, focusing on Kubernetes and messaging platforms.
Badges
Area of Expertise
Topics
The Cost of Apache Kafka
Strimzi makes it easy to run Apache Kafka on Kubernetes. But persistent volumes and high-throughput cross-zone traffic can quietly become major cost drivers. In this talk, we’ll show how you can use OpenCost and NetObserve to build an open source, cost-aware view of our Kafka infrastructure — and how we tuned it to be leaner and smarter. While OpenCost doesn’t track network costs out of the box, we’ll show how to extend it with NetObserve data to estimate cross-zone network costs, attributing them directly to the pods generating the traffic. OpenCost can also help us to monitor compute and storage costs. We’ll share an open-source-driven playbook for cost-effective Apache Kafka on Kubernetes, informed by real-world experience and actionable strategies, without relying on proprietary tools or commercial observability platforms.
Rethinking Observability: Decoupling Control and Data Planes
As cloud-native adoption grows, so do telemetry volumes, and with them, observability bills. High-cardinality metrics, distributed traces, and logs are essential for modern applications, but they come at a steep cost. Centralized vendor platforms exacerbate this with usage-based billing, while open source stacks, though cost-effective, often lack a unified control plane.
What if we decouple the control plane from the data plane?
In this talk, we’ll explore Federated Control Planes as an architectural pattern for building cost-effective observability with on-demand telemetry retrieval. We’ll show how OpenTelemetry plays a central role in making this possible, and share the challenges, trade-offs, and lessons learned from real-world implementations.
Getting Started with OpenTelemetry: The Future of Observability
OpenTelemetry is revolutionizing observability, offering a unified framework for tracing, metrics, and logs. In this session, we’ll cover the fundamentals of OpenTelemetry, key observability patterns, and best practices for instrumenting applications. We’ll walk through how OpenTelemetry helps track requests across services, measure performance, and diagnose issues effectively. To bring these concepts to life, we’ll showcase a live demo, demonstrating how to set up OpenTelemetry and use it to gain real-time insights into system behavior. Whether you're new to OpenTelemetry or looking to integrate it into your stack, this session will provide a solid foundation to get started and unlock the full potential of modern observability.
Cutting Kubernetes Costs: Strategies and Tools for Efficient Cluster Management
As Kubernetes adoption grows, so do its operational costs. Without proper cost management, resource waste can spiral out of control. This session covers Kubernetes cost fundamentals, exploring how workloads consume resources and where inefficiencies arise. We’ll dive into rightsizing, autoscaling optimization, and OpenCost for better visibility. Attendees will gain practical insights to cut unnecessary spend while maintaining performance. Join us to master Kubernetes cost efficiency!
3 Tips to Take Control of Your Telemetry Data with OTel
OpenTelemetry gives you freedom to take control of your telemetry data it. In this fast-paced 10-minute demo, we’ll show three practical ways to take control of your telemetry data to improve signal, reduce noise and decrease your cost.
Whether you’re just getting started with OTel or looking to fine-tune your observability pipeline, you’ll walk away with proven techniques you can implement today.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top