Real-Time Topology to Connect Platform Engineering & SRE
summary
Claims that feeding a continuously autodiscovered runtime topology into an internal developer portal closes the loop between config and reality, using BPF and machine learning to map live Kubernetes relations and third‑party dependencies. Demonstrates an AIOps pipeline that auto‑discovers topology, flags surprising resource shifts (example: a new service spikes CPU at peak), verifies deployments in real time, and pushes adjusted guardrails back into the IDP, while warning about maintenance cost and the need for continuous autodiscovery.
AI-written summary · about
chapters (11)
- 0:00intro
- 0:14overview
- 2:06understanding internal developer portals (idps)
- 3:26challenges in idp implementation
- 5:16bridging the gap with real-time topology
- 5:44cybersecurity analogies and lessons
- 7:52benefits of real-time topology in idps
- 14:38practical use cases and examples
- 23:14common pitfalls and how to avoid them
- 25:26tips for getting started
- 29:00conclusion and final thoughts
related talks
26mReal-Time Application Maps for Proactive and Actionable Visibility
46mObservability in the real world: Improve systems w/iterative approach
42mAIOps for Distributed Environments - Deep Dive
23mFrom Firefighting to Zero-Touch
37mMonitoring Service Architecture and Health with BPF
41mCustomer Centric Observability at Scale Leveraging AIOps & OpenTelemetry
53mFull-stack observability in a world that prioritizes applications (PRT201)
15mPlatforming the Platform: Running Our IDP With Our IDP