Recently, the reliability of some developer platforms has been impacted by out-of-memory errors, causing pods...

The NKPA course details NKP’s monitoring stack for collecting and analyzing metrics across attached and managed clusters, which is critical for troubleshooting issues like out-of-memory errors. The engineer needs to collect centralized metrics and set up alerting to identify performance issues. The correct approach is to use Prometheus for metrics collection and Thanos Query for alerting (Option C).

Utilize Prometheus to collect and present centralized metrics on NKP attached and managed clusters: Prometheus is NKP’s default metrics collection tool, deployed as a platform application. It scrapes metrics (e.g., CPU, memory usage) from nodes, pods, and services across all clusters in a workspace, centralizing them for analysis. The Nutanix Cloud Native (NCP-CN) 6.10 Study Guide states: “Prometheus is used in NKP to collect and present centralized metrics from attached and managed clusters, enabling detailed performance analysis.” This is ideal for identifying differences in memory usage causing the out-of-memory errors.

Configure Thanos Query to monitor and send out alerts based on the metric thresholds that have been configured within it: Thanos Query, part of NKP’s monitoring stack, aggregates Prometheus metrics across clusters and provides a unified query interface. It also integrates with Alertmanager to send alerts based on configured thresholds (e.g., memory usage > 90%). The NKPA course notes: “Use Thanos Query to monitor metrics across NKP clusters and configure alerts with Alertmanager to notify engineers of performance issues like out-of-memory errors.” This setup helps the engineer proactively address the issue.

Incorrect Options:

A & B. Utilize Fluentd and Fluent Bit: Fluentd and Fluent Bit are logging tools in NKP, used for collecting logs, not metrics. Metrics collection requires Prometheus.

D. Utilize Thanos to collect and present centralized metrics: Thanos does not collect metrics directly; it extends Prometheus by providing long-term storage and query aggregation. Prometheus is the primary collection tool.