Fast and Furious Multi-Cloud Connectivity
summary
Company implements multi-cloud across AWS GCP and Azure using Kubernetes, Terraform and Cilium, finds VPN unreliable for latency and routing spikes so uses direct data‑center connectivity with redundant links instead of Megaport; setup takes 2–3 months but yields p99 latency halves with direct connectivity and steadier bandwidth. Tests use iperf/netperf and HTTP benchmarks; cloud‑native CNIs match Cilium for throughput, but Cilium lacks intra‑node encryption and its global services use only round‑robin, causing affinity and security gaps.
AI-written summary · about
chapters (5)
- 0:00Start
- 3:33Challenges of payments processing
- 18:24Requirements
- 21:55Virtual Private Network
- 34:13Performance Tests
related talks
47mFast and Furious Multi-Cloud Connectivity
24mConfluent's Multi-Cloud Journey to Cilium: Pitfalls and Lessons Learned
27mMeshing It up Securely: How Roche Manages Network Connectivity for...
09mCilium on Azure: Most Scalable and Performant Implementation in...
27mChoosing Cloud Native Technologies for the Journey to Multi-cloud
31mCilium Mesh
05mLightning Talk: Unveiling 5 Essential Insights from Cilium Implementation in a...
18meBPF & Cillium
