Exam-style
A production Bedrock application needs alerting when invocation latency or throttling climbs, and the compliance team separately wants to inspect the exact prompt and response behind a specific customer complaint. What combination covers both needs?
Reveal the answer
B. CloudWatch metrics and alarms for latency and throttling, plus model invocation logging for the content
The two questions need two mechanisms. CloudWatch publishes Bedrock invocation counts, latency, throttles and token usage, which is what you alarm on, while model invocation logging is the feature that captures prompt and response payloads to S3 or CloudWatch Logs. CloudTrail records who called the API and what configuration changed, not the content of an invocation or its latency distribution. X-Ray and SDK debug logging can be made to carry payloads, but that is a bespoke pipeline putting sensitive content wherever a developer chose. Config reports resource compliance and has no view of runtime invocations.
Q. What gives you operational visibility into a production Bedrock app?
A. CloudWatch metrics (invocation counts, latency, throttling, token usage) and alarms, plus model invocation logging for prompt and response inspection. CloudTrail covers configuration changes.
Why? Ops metrics and content records answer different questions; you want both.