A customer request passes through applications in one cloud and a database service in another. Which observability design is most useful for diagnosing end-to-end failures?

1. Disable all provider logs to reduce the number of consoles
2. Assume every provider automatically shares all operational data without configuration
3. Correlate relevant logs, metrics, request identifiers, and timing across the participating services
4. Inspect only one unrelated infrastructure graph