Questão de System Administration #019e3a4f

Metrics are the most suitable observability data stream for plotting resource consumption and predicting future resource exhaustion. Metrics are numerical measurements collected over time, such as CPU usage, memory utilization, disk I/O, and network throughput. They are typically stored in time-series databases (e.g., Prometheus, Graphite) and can be aggregated, averaged, and analyzed using mathematical models to detect trends and forecast future values. Tools like PromQL or Grafana allow querying and visualizing metric data, enabling capacity planning and proactive alerts.

How it Works:

• Metrics are collected at regular intervals (e.g., every 15 seconds) from system components via exporters (e.g., node_exporter for Linux).
• They are stored with timestamps, forming a time series.
• Predictive algorithms (e.g., linear regression, Holt-Winters) can be applied to historical metric data to estimate when a resource (e.g., disk space) will be exhausted.
• This allows administrators to set thresholds and receive warnings before critical failures occur.

Why the other options are incorrect:

• stdout is a standard output stream for command-line programs. It is not a structured observability data stream; it contains arbitrary text output and is not designed for time-series analysis or prediction.
• Traces represent the path of a single request through a distributed system (e.g., in microservices). They are useful for debugging latency and errors but are not aggregated numerical data suitable for plotting resource consumption trends.
• Logs are unstructured or semi-structured text records of events. While logs can indicate resource issues (e.g., "disk full"), they are not numerical time-series data and are inefficient for plotting and predictive modeling due to their high volume and lack of numeric aggregation.