Cloud Report.pdf

The analysis covers the period from October 1, 2019, to October 31, 2022.

Key Insights:

1. Performance issues are not rare: Cloud providers experience challenges and performance missteps, and enterprises should expect and plan for issues of all sizes and frequencies.
2. Cloud providers show preferences in managing their networks: Cloud providers have priorities and preferences in fixing issues and optimizing performance, which can impact organizations.
3. No steady state in the cloud: Public cloud connectivity architectures continuously evolve, and network performance can vary substantially over time and in different regions.

Summary of Findings:

1. Connectivity architectures: Cloud providers' decisions on advertising service endpoints, prioritizing optimizations, and leveraging shared infrastructure can impact customers' performance and operations.
2. End-user performance: End-user performance varies by region, with Asia and Oceania experiencing higher latency and packet loss compared to other regions.
3. Cloud provider network preferences: AWS, Azure, and Google Cloud have different network preferences, with AWS leveraging the Internet more extensively, while Azure and Google Cloud use their backbones more frequently.

Recommendations:

1. Monitor routes and performance: Enterprises should monitor their routes and performance to ensure that network events or traffic engineering activities do not impact operations or business continuity.
2. Understand cloud provider network preferences: Organizations should understand cloud providers' network preferences and prioritizations to ensure they are not impacted.
3. Continuously evaluate and optimize: Enterprises should continuously evaluate and optimize their cloud management strategy to account for the dynamic nature of the cloud.