Custom Log Based Metrics Counter Distribution Design

Worked example. The correct answer is already marked and every option is explained below, so there is nothing to select here. To answer questions yourself, start the free trial.

A DevOps team manages a microservices application running on Google Cloud where services emit structured JSON logs to Cloud Logging. The logs contain a response status code (jsonPayload.status), the name of the service (jsonPayload.service_name), and the execution duration in milliseconds (jsonPayload.latency_ms).

The team needs to implement observability solutions in Cloud Monitoring to fulfill two requirements:

Monitor latency percentiles (such as 95th and 99th percentiles) over time and trigger alerts when tail latency degrades.
Track the rate of HTTP 5xx server errors broken down by microservice and trigger alerts on sudden error rate spikes.

Which custom log-based metric designs should the team configure in Cloud Logging?

Create a distribution log-based metric extracting jsonPayload.latency_ms to record latency into histogram buckets, and create a counter log-based metric filtering for jsonPayload.status >= 500 with a custom label extracting jsonPayload.service_name.

Create a counter log-based metric extracting jsonPayload.latency_ms with configured histogram buckets, and create a distribution log-based metric filtering for jsonPayload.status >= 500.

Create a single distribution log-based metric that extracts both jsonPayload.latency_ms and jsonPayload.status, and use Cloud Monitoring metric-threshold filters to count 5xx status codes.

Create a boolean log-based metric for latency percentiles, and configure a log-based alerting policy directly in Logs Explorer to monitor 5xx error rate trends.

Professional Cloud DevOps Engineer

Custom Log Based Metrics Counter Distribution Design

Related questions