Production Value: Networking is the vascular system of cloud-native architectures. 90% of SRE debug calls during outages trace back to DNS routing limits, network policy blocks, subnet space exhaustion, or asymmetrical route paths. Master these rules to isolate failure domains instantly.

Day-2 Support Tips:
• Prerequisite Validation: When starting a debug flow from a middle tier, check DNS first (dig +trace), confirm firewall policies next (NACLs/SG), and trace pod CNI paths with eBPF metrics.
• DNS Latency: Large clusters can overload CoreDNS. Use NodeLocal DNSCache to cache DNS responses directly on worker node loopbacks, decreasing resolution latency from ~15ms to <1ms.
• Subnet Reservation: Always allocate subnets with growth in mind (at least /20 for K8s nodes). Pod overlays consume IPs rapidly, and changing subnet sizes in AWS after provisioning requires recreating resources.
• L4 vs L7 routing: Use NLB (L4) for high-performance TCP traffic or raw protocol streams. Use ALB (L7) when header-based filtering, SSL termination, or context-routing is required.
• Firewall Filtering: Place stateful security groups as close to target instances as possible. NACLs are stateless and should be reserved for broad traffic blocking or boundary isolation.

🔍 Phase 1: Identify Knowledge Level