Search papers, labs, and topics across Lattice.
This paper provides a comprehensive empirical analysis of Google Cross-Cloud Interconnect (CCI) and its alternatives, focusing on the challenges of cost optimization due to unpredictable traffic patterns and high fixed leasing fees. The authors introduce ToggleCCI, a dynamic cost-optimization algorithm that intelligently switches between VPN and CCI based on observed cost trends, effectively managing provisioning delays and demand uncertainty. The results demonstrate that ToggleCCI achieves asymptotic optimality and significantly outperforms static policies in real-world scenarios, yielding substantial cost savings.
ToggleCCI adapts to unpredictable traffic patterns, delivering significant cost savings by dynamically switching between VPN and CCI based on real-time cost trends.
New services such as Google Cross-Cloud Interconnect (CCI) address the rise in fast and large-scale cross-cloud data transfers. CCI offers dedicated high-throughput links with low per-GB transfer costs, but also involves high fixed leasing fees and multi-day provisioning delays. This combination makes cost optimization difficult because traffic patterns are unpredictable. This paper presents the first comprehensive study of CCI-like services. We begin with an empirical characterization of CCI and its alternatives using direct measurements across AWS-GCP interconnects. We then introduce ToggleCCI, a new dynamic cost-optimization algorithm designed to handle provisioning delays and uncertainty in future demand. ToggleCCI adapts by switching between VPN and CCI based on cost trends observed over a sliding time window. We prove that ToggleCCI achieves asymptotic optimality under sustained high-demand or low-demand regimes. Finally, using real-world traffic traces, we show that ToggleCCI consistently tracks the best static policy for each scenario and delivers substantial cost savings.
BAIR