Reliability To the Edge: Telco Teaching Cloud Native
summary
Telco practices mapped to cloud-native edge focus on measurable guarantees: 99.999% availability (about five minutes/year) and zero‑trust per NIST SP 800‑207. Concrete stack choices: Kubernetes plus a UDSF (session store, e.g., Redis) for session continuity, service mesh and policy points for microsegmentation, L4/eBPF to offload L7, chaos engineering for recovery, and twin hardware plus multiple power feeds at far edge. An IIT Bombay paper reports near‑zero session loss using UDSF but a ~25% throughput penalty and requires 4+ hardware replicas.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:04Defining Telco reliability
- 2:48How Telco achieves 99.999% availability
- 4:54UDSF for session resilience
- 6:48Edge reliability challenges
- 8:575G fault tolerance study
- 11:30Cost versus critical services
- 12:32Zero trust in regulated networks
- 14:46Zero trust culture challenges
- 15:09Assessing reliability needs
- 16:13Reliability beyond tools
- 19:28Q&A: RAN reliability
- 21:47Q&A: Sovereignty resiliency
- 23:33Q&A: Hardware costs
related talks
40mSecuring Connections: Defending Telco Workloads in the Cloud Era
41mCloud Computing: The Edge talks - long road into the Cloud
21mUnpacking “Edge” In Verizon 5GEdge
37mEdge-Native Application Principles: Taking Your App Beyond the Cloud
07mScalable Edge Data Centers: Current Challenges and How to …
30mLessons Learnt and Thoughts on Telco Cloud Native Evolution and Open Source Practices
37mCloud Native Is Good, but How to Apply It in Telecom Networks?
21mKeynote: Cloud Native: Essential for 5G