Prometheus Certified Associate (PCA) Practice Exam

What the Prometheus Certified Associate (PCA) exam covers

• Observability Concepts85 questions
• Prometheus Fundamentals168 questions
• PromQL156 questions
• Instrumentation and Exporters153 questions
• Alerting and Dashboards136 questions

Free Prometheus Certified Associate (PCA) sample questions

A sample of 10 questions with answers and explanations. Sign up free to practice all 698.

1. Question 1Observability Concepts

   What type of monitoring system is Prometheus?

   • AA metrics-based, time-series monitoring and alerting system that pulls (scrapes) metricsCorrect
   • BA log aggregation and full-text search platform that indexes application log lines
   • CA distributed tracing backend that stores request spans across microservices
   • DAn event-streaming message broker that buffers telemetry between producers and consumers
   ✓ Correct answer: A

   Prometheus is a metrics-based, time-series monitoring system that pulls metrics via HTTP scraping. Prometheus provides foundational observability for modern cloud infrastructure. Understanding these concepts is essential for effective monitoring strategy.

   Why the other options are wrong

   • BThat describes log-search systems like Loki or Elasticsearch, not Prometheus, which stores numeric time series.
   • CStoring request spans describes a tracing backend such as Jaeger or Tempo, not Prometheus.
   • DBuffering telemetry as a message broker describes Kafka-style systems, not a metrics monitoring server.
2. Question 2Prometheus FundamentalsSelect all that apply

   You are integrating two regional Prometheus servers into a single global view for cross-region dashboards and long-term retention. Which TWO approaches are appropriate? (Choose TWO)

   • AConfigure remote_write from each regional Prometheus into Thanos/Mimir for a global, long-term storeCorrect
   • BEnable DNS zone transfers between the two Prometheus servers
   • CUse a hierarchical federation server that scrapes aggregated series from each regional Prometheus via /federateCorrect
   • DPoint Alertmanager at both servers so it merges their TSDBs
   ✓ Correct answer: A, C

   use a hierarchical federation server that scrapes aggregated series from each regional Prometheus via /federate Both configure remote_write from each regional Prometheus into Thanos/Mimir for a global, long-term store and use a hierarchical federation server that scrapes aggregated series from each regional Prometheus via /federate are essential components within the Prometheus Fundamentals domain. These features work together to provide comprehensive functionality required for effective platform implementation. Understanding how these capabilities integrate is critical for certification-level knowledge.

   Why the other options are wrong

   • BEnable DNS zone transfers between the two Prometheus servers is incorrect because DNS transfers have no relation to metrics collection.
   • DPoint Alertmanager at both servers so it merges their TSDBs is incorrect because this does not provide the specific technical capability required by the question.
3. Question 3Instrumentation and Exporters

   Which flag changes the address and port that node_exporter binds its metrics endpoint to?

   • A--web.address=:9200
   • B--web.listen-address=:9200Correct
   • C--listen=:9200
   • D--bind-address=:9200
   ✓ Correct answer: B

   Quantitative measurements of system behavior collected and analyzed over time. Prometheus provides foundational observability for modern cloud infrastructure. Understanding these concepts is essential for effective monitoring strategy.

   Why the other options are wrong

   • A--web.address=:9200 is incorrect because this does not provide the specific technical capability required by the question.
   • C--listen=:9200 is incorrect because this does not provide the specific technical capability required by the question.
   • D--bind-address=:9200 is incorrect because this does not provide the specific technical capability required by the question.
4. Question 4Instrumentation and Exporters

   An exporter exposes thousands of rarely-used metrics, inflating scrape size and storage cost. You want Prometheus to keep only a handful of metric names from that job. Which configuration is the correct, lowest-overhead way?

   • AA metric_relabel_configs keep action matching __name__ against the allowed metric namesCorrect
   • BA recording rule that copies only the wanted metrics into new series
   • CLower the exporter's log verbosity so it emits fewer metric lines
   • DA relabel_configs drop action applied to the target's __address__ label
   ✓ Correct answer: A

   Prometheus is a metrics-based, time-series monitoring system that pulls metrics via HTTP scraping. This pull-based approach enables monitoring without requiring targets to be configured with collector addresses. Targets simply expose a /metrics endpoint that Prometheus periodically requests.

   Why the other options are wrong

   • BRecording rules add series on top of the ingested ones; they do not stop the unwanted metrics from being stored.
   • CLog level controls exporter logging, not which metrics are exposed or ingested.
   • DDropping on __address__ removes whole targets, not specific metric names, and runs before samples exist.
5. Question 5Alerting and Dashboards

   You need an alert that detects when Prometheus stops being able to scrape a critical target, which is often the first sign of a security incident or outage. Which expression is the canonical basis for this alert?

   • Arate(scrape_samples_scraped[5m]) > 0
   • Bhistogram_quantile(0.9, up)
   • Cabsent(scrape_duration_seconds) only
   • Dup{job="critical"} == 0 sustained with a for clauseCorrect
   ✓ Correct answer: D

   Prometheus is a metrics-based, time-series monitoring system that pulls metrics via HTTP scraping. This pull-based approach enables monitoring without requiring targets to be configured with collector addresses. Targets simply expose a /metrics endpoint that Prometheus periodically requests.

   Why the other options are wrong

   • Arate(scrape_samples_scraped[5m]) > 0 is incorrect because this does not provide the specific technical capability required by the question.
   • Bhistogram_quantile(0.9, up) is incorrect because this does not provide the specific technical capability required by the question.
   • Cabsent(scrape_duration_seconds) only is incorrect because this does not provide the specific technical capability required by the question.
6. Question 6PromQL

   Which PromQL function smooths a gauge time series by computing the exponentially weighted moving average (EWMA) over a range?

   • Aavg_over_time()
   • Bholt_winters()Correct
   • Cderiv()
   • Dsmooth()
   ✓ Correct answer: B

   holt_winters(v range-vector, sf, tf) applies double exponential smoothing to a range vector, where sf is the smoothing factor (0-1) and tf is the trend factor (0-1). Lower sf weights older data more, producing a smoother line. It is designed for gauges and is useful for smoothing noisy signals or building trend-sensitive alerts.

   Why the other options are wrong

   • Aavg_over_time() computes a simple arithmetic mean over the range window - it does not apply exponential weighting or track trends.
   • Cderiv() calculates the per-second derivative of a gauge using linear regression - it returns a rate of change, not a smoothed value.
   • Dsmooth() is not a valid PromQL function - no such function exists in the Prometheus query language.
7. Question 7Alerting and Dashboards

   Which promtool command validates the syntax of an Alertmanager configuration file before deploying it?

   • APromtool check config alertmanager.yml
   • BPromtool check rules alertmanager.yml
   • CAmtool check-config alertmanager.ymlCorrect
   • DPromtool test rules alertmanager.yml
   ✓ Correct answer: C

   The amtool binary (shipped with Alertmanager) provides the 'check-config' subcommand to parse and validate an alertmanager.yml file without starting the Alertmanager process. It reports YAML syntax errors, unknown fields, and structural issues. promtool is for Prometheus configuration and rules files, not Alertmanager configuration.

   Why the other options are wrong

   • Apromtool check config validates prometheus.yml (the Prometheus server configuration), not alertmanager.yml.
   • Bpromtool check rules validates Prometheus alerting and recording rule files; alertmanager.yml is not a rule file and cannot be validated by this subcommand.
   • Dpromtool test rules runs unit tests defined in test YAML files against Prometheus rules; it is unrelated to Alertmanager configuration validation.
8. Question 8Prometheus Fundamentals

   In the Prometheus text exposition format, what does a line beginning with `# HELP` provide?

   • AA single-line, human-readable description of the metricCorrect
   • BThe declared data type of the metric (counter, gauge, etc.)
   • CThe current timestamp for the following sample
   • DA list of all label values that the metric may emit
   ✓ Correct answer: A

   In the exposition format, `# HELP <metric> <text>` documents what the metric measures and is shown in tooling such as the expression browser. It is purely descriptive metadata and does not affect ingestion. The `# TYPE` comment, not `# HELP`, declares the metric type.

   Why the other options are wrong

   • BDeclaring the metric type is done by the `# TYPE` line, not `# HELP`.
   • CTimestamps are optional fields appended after a sample value, not provided by `# HELP`.
   • DThe format does not enumerate possible label values; labels appear inline on each sample.
9. Question 9PromQL

   Which PromQL function returns the per-second instantaneous rate of a counter using only the two most recent samples in the range vector?

   • Arate()
   • Birate()Correct
   • Cincrease()
   • Ddelta()
   ✓ Correct answer: B

   irate() looks only at the final two data points in the supplied range vector, making it highly responsive to brief spikes but jagged on dashboards. It is counter-aware and corrects for resets like rate().

   Why the other options are wrong

   • Arate() averages the per-second increase across all samples in the entire window, not just the last two.
   • Cincrease() returns the total increase over the window (not per-second) and is essentially rate() multiplied by the window seconds.
   • Ddelta() is intended for gauges and computes the difference between first and last samples without counter-reset handling or per-second scaling.
10. Question 10Instrumentation and Exporters

    You scrape a target and notice every series gained an instance label equal to the address from __address__. What is responsible for setting the instance label by default?

    • AAn explicit relabel_config you must always write
    • BDefault relabeling that copies __address__ to instance if it is unsetCorrect
    • CThe exporter, which injects instance into each sample
    • DThe Pushgateway, which adds instance on ingestion
    ✓ Correct answer: B

    After service discovery and relabeling, if no instance label has been assigned, Prometheus defaults it to the value of __address__ (the host:port being scraped). This guarantees each scraped target has a distinguishing instance label.

    Why the other options are wrong

    • AYou do not need to write it; it is applied automatically by default.
    • CThe exporter does not set instance; Prometheus assigns it during relabeling.
    • DThe Pushgateway is unrelated to default scrape relabeling.

Related Cloud Native resources

• Prometheus Certified Associate (PCA) study guideKey concepts
• Cloud Native practice examsAll Cloud Native
• Certification pathWhere this fits
• Certification exam guides & tipsBlog
• Plans & pricingFree & paid
• KCSA practice examRelated
• CKS practice examRelated
• CKA practice examRelated

Prometheus Certified Associate (PCA) practice exam FAQ