N Is For Networking N4N044: Redundancy Vs. High Availability Part 2 – HA Networking Isn’t Free
12 snips
Dec 4, 2025 Dive into the nuances of redundancy versus high availability in networking. The hosts clarify how high availability serves as a design goal while redundancy is a tool to achieve it. Learn about rapid failure detection with BFD, the pros and cons of active-active versus active-standby setups, and how business recovery expectations shape design decisions. They explore the role of AI in proactive failure detection and analyze the cost of outages, wrapping up with insights on balancing design trade-offs and budget constraints.
AI Snips
Chapters
Transcript
Episode notes
Apply QoS To Enable Graceful Degradation
- Implement QoS to prioritize critical traffic and make congestion degrade gracefully.
- Protect voice and video with low-latency queues to preserve user experience.
ECMP Enables Partial Failure Tolerance
- ECMP provides graceful degradation by distributing load across parallel links.
- Losing one ECMP member reduces capacity but preserves connectivity.
Proactive Monitoring Predicts Failures
- Proactive detection predicts failures using telemetry and historical patterns.
- AI can surface subtle signals like optics or ASIC errors before they become outages.