Multi Cloud Observability for 5G
summary
Argues that sidecar/service-mesh tracing adds unacceptable overhead at telco scale and instead demonstrates host-level collection: daemonsets plus an agentless OVS pull from virtual switches feeding a central OpenShift hub. Implements a working pipeline with an NetObserve operator tech preview pointing clusters (AWS and VMware 5G cores) at a Grafana Loki backend, visualizing Open5G Core metrics and traces from a single pane. Publishes code and reproducible configs in a GitHub repo and Medium article.
AI-written summary · about
chapters (17)
- 0:00Introduction to speakers
- 1:17Speaker backgrounds
- 2:11Observability challenges
- 3:23Fatih's professional journey
- 4:355G network complexity
- 6:48Simplifying complex systems
- 8:26Zero touch provisioning
- 10:02Logs, metrics, and traces
- 12:04Service mesh overhead
- 14:07Daemon set approach
- 15:48StackRox security
- 17:05Net observer operator
- 19:10Distributed 5G architecture
- 21:47Network slicing and SON
- 23:46Hyperscaler infrastructure
- 25:11Detailed article and repo
- 26:40Q&A session
related talks
27mObservability Brings Clarity in 5G World
22mHarnessing Observability for 5G Performance: eBPF and OpenTelemetry I...
31mCloud Native Observability
25mOpen source observability for private cloud: mission impossible or not?
26mObservability & Monitoring in Cloud Native Telco World – Tying It Together
05mLightning Talk: Retire the Sidecar: Why Observability Needs To Ride up Front
28mUnlocking Multi-Cloud Observability: The Case Study of...
01h 29mTutorial: A Hitchhiker's Guide to CNCF/OSS Observability So...