PracticeVideosBlogFlashcardsEspañol
All Practice Exams

100+ Free Prometheus Certified Associate Practice Questions

Prepare for the Prometheus Certified Associate (PCA) exam with instant access — no signup required.

✓ No registration✓ No credit card✓ No hidden fees✓ Start practicing immediately
100+ Questions
100% Free

Loading practice questions...

Same family resources

Explore More Linux Foundation Certifications

Continue into nearby exams from the same family. Each card keeps practice questions, study guides, flashcards, videos, and articles in one place.

Linux Foundation Certified System Administrator (LFCS)

Practice Questions

Besu Certified Professional (BCP)

Practice Questions

Certified Backstage Associate (CBA)

Practice Questions

Certified Open Source Developer for Enterprise (CODE)

Practice Questions

Cilium Certified Associate (CCA)

Practice Questions

FCFP FINOS Certified FDC3 Practitioner

Practice Questions
2026 Statistics

Key Facts: Prometheus Certified Associate Exam

$250

Exam Fee (includes one free retake)

Linux Foundation / CNCF

90 minutes

Exam Duration

Linux Foundation

28%

PromQL, the largest exam domain

CNCF PCA curriculum

2 years

Certification Validity

Linux Foundation

12 months

Exam Eligibility Window

Linux Foundation

No prerequisites

Eligibility Requirement

Linux Foundation

The Prometheus Certified Associate (PCA) is a CNCF/Linux Foundation pre-professional exam that is online, proctored, and multiple-choice, lasts 90 minutes, and costs $250 with one free retake. The Linux Foundation does not publish a fixed question count or an official passing score (it is commonly estimated near 66%). The five domains are Observability Concepts (18%), Prometheus Fundamentals (20%), PromQL (28%), Instrumentation and Exporters (16%), and Alerting and Dashboarding (18%). There are no prerequisites, the purchase includes a 12-month eligibility window, and the certification is valid for two years.

Sample Prometheus Certified Associate Practice Questions

Try these sample questions to test your Prometheus Certified Associate exam readiness. Each question includes a detailed explanation. Start the interactive quiz above for the full 100+ question experience with AI tutoring.

1In the context of observability, what are the three primary signal types (often called the 'three pillars') that systems emit to describe their internal state?
A.Metrics, logs, and traces
B.Alerts, dashboards, and reports
C.Counters, gauges, and histograms
D.Scrapes, queries, and rules
Explanation: Observability is commonly described through three pillars: metrics (numeric measurements over time), logs (timestamped records of discrete events), and traces (the path of a request across services). Prometheus focuses primarily on the metrics pillar.
2Which statement best describes why Prometheus uses a pull-based model for collecting metrics rather than a push-based model?
A.Pull requires no network configuration on the targets
B.Pull guarantees metrics are never lost during a network partition
C.The Prometheus server scrapes targets over HTTP, which centralizes control over scrape timing, makes target health visible via the up metric, and simplifies running multiple Prometheis
D.Pull lets applications send metrics only when a value changes, reducing bandwidth
Explanation: Prometheus pulls (scrapes) metrics from targets over HTTP. This centralizes scrape scheduling, lets Prometheus detect a down target by exposing the synthetic up metric, and makes it trivial to point several Prometheus servers at the same targets. Push gateways exist only for short-lived batch jobs.
3A team defines a target of '99.9% of HTTP requests succeed over a rolling 30-day window'. In SRE terminology, what is this statement an example of?
A.A Service Level Indicator (SLI)
B.An error budget policy
C.A Service Level Agreement (SLA)
D.A Service Level Objective (SLO)
Explanation: An SLO is a target value or range for an SLI over a time window, such as 99.9% success over 30 days. The SLI is the underlying measured ratio (successful requests / total requests), and an SLA is the externally contracted commitment with consequences.
4What is the relationship between an SLO and an error budget?
A.The error budget is always 100% minus the SLA penalty
B.The error budget is the portion of failures allowed by the SLO over the measurement window
C.The error budget equals the SLO target percentage
D.The error budget is the number of alerts a service may fire per day
Explanation: An error budget is the allowable amount of unreliability implied by an SLO. For a 99.9% availability SLO, the error budget is the remaining 0.1% of requests (or time) that may fail in the window before the objective is breached.
5In distributed tracing, what is the relationship between a trace and a span?
A.A span contains many traces, one per service
B.A trace represents one request's journey and is composed of one or more spans, each representing a unit of work
C.A trace and a span are interchangeable names for the same object
D.A span is a numeric metric while a trace is a log line
Explanation: A trace represents the end-to-end path of a single request through a system. It is made up of spans, where each span captures a single operation or unit of work (such as a function call or an RPC) with a start time, duration, and contextual attributes.
6Which of the following workloads is the Prometheus Pushgateway specifically designed to support, given that Prometheus normally pulls metrics?
A.High-cardinality streaming metrics from a web frontend
B.Short-lived or batch jobs that may finish before Prometheus can scrape them
C.Long-running services behind a load balancer
D.Storing historical metrics for years of retention
Explanation: The Pushgateway is an intermediary for ephemeral and batch jobs that exit before Prometheus would scrape them. The job pushes its metrics to the Pushgateway, which Prometheus then scrapes. It is explicitly not a general-purpose way to convert Prometheus into a push system.
7Which signal type is BEST suited for efficiently answering 'how many requests per second is this service handling, aggregated across all instances?' at low storage cost?
A.Traces
B.Logs
C.Metrics
D.Profiles
Explanation: Metrics are pre-aggregated numeric time series, making them cheap to store and fast to aggregate across instances. Computing request rate from logs or traces would require scanning far more data. This efficiency is a core reason Prometheus is metrics-based.
8A monitoring team wants Prometheus to automatically find new pods to scrape as they are created and destroyed in a Kubernetes cluster. Which Prometheus mechanism provides this capability?
A.Recording rules
B.The Pushgateway
C.Service discovery
D.Remote write
Explanation: Service discovery lets Prometheus query a dynamic source (such as the Kubernetes API, Consul, or EC2) to learn which targets currently exist, instead of listing them statically. As pods come and go, kubernetes_sd_configs keeps the scrape target list current.
9What is the key difference between monitoring and observability as the terms are typically used?
A.Monitoring is only for hardware, observability only for software
B.Observability requires no instrumentation, monitoring does
C.Monitoring watches predefined known failure conditions, while observability emphasizes exploring a system's internal state to answer questions you did not anticipate
D.They are the same thing with no practical difference
Explanation: Monitoring traditionally tracks known indicators and alerts on predefined thresholds (the 'known unknowns'). Observability is the broader property of being able to ask new, unanticipated questions about a system from its telemetry (the 'unknown unknowns'), typically by combining metrics, logs, and traces.
10When Prometheus scrapes a target and the scrape succeeds, what value does the synthetic up metric for that target receive?
A.0
B.The scrape duration in seconds
C.The HTTP status code
D.1
Explanation: For every scrape, Prometheus generates a synthetic up sample: 1 if the target was reachable and the scrape succeeded, and 0 if it failed. Querying up == 0 is a standard way to alert on down targets.

About the Prometheus Certified Associate Exam

The Prometheus Certified Associate (PCA) is a pre-professional certification from the Cloud Native Computing Foundation and the Linux Foundation that demonstrates foundational knowledge of observability and practical skills with Prometheus, the open source systems monitoring and alerting toolkit. The exam is online, proctored, and multiple-choice, runs for 90 minutes, and costs $250 with one free retake included. Its blueprint spans five domains: Observability Concepts (18%), Prometheus Fundamentals (20%), PromQL (28%), Instrumentation and Exporters (16%), and Alerting and Dashboarding (18%). Candidates are tested on metrics, the pull model and service discovery, the data model and exposition format, writing PromQL (rate, aggregation, histogram_quantile), instrumenting code with client libraries and exporters, and configuring alerting rules with Alertmanager grouping, inhibition, and silences. The credential is valid for two years.

Assessment

Question count not published by the exam provider

Time Limit

90 minutes

Passing Score

Not officially published (commonly estimated around 66%)

Exam Fee

$250 (includes one free retake) (CNCF / The Linux Foundation)

Prometheus Certified Associate Exam Content Outline

18%

Observability Concepts

Understand metrics, logs and events, tracing and spans, the push vs pull collection model, service discovery sources such as Kubernetes and Consul, and the definitions of SLOs, SLAs, SLIs, and error budgets.

20%

Prometheus Fundamentals

Know the Prometheus server and TSDB architecture, scrape configuration (job_name, scrape_interval, relabeling, the up metric), the label-based data model, the text exposition format with HELP and TYPE, and Prometheus's design limitations around accuracy and long-term storage.

28%

PromQL

Write queries that select data with label matchers, compute counter rates with rate, irate, and increase, derive gauge change with deriv and delta, aggregate over time and over dimensions with by and without, apply binary and set operators, estimate quantiles with histogram_quantile, and work with timestamp metrics.

16%

Instrumentation and Exporters

Use client libraries (Go, Java, Python, Ruby) to instrument code, choose between counter, gauge, histogram, and summary, deploy exporters such as node_exporter and blackbox_exporter, and follow naming conventions including base units and the _total suffix while avoiding high-cardinality labels.

18%

Alerting and Dashboarding

Build dashboards in Grafana with PromQL, define recording and alerting rules with the for clause and pending/firing states, and operate Alertmanager for grouping, inhibition, silences, receivers, and routing trees.

How to Pass the Prometheus Certified Associate Exam

What You Need to Know

  • Passing score: Not officially published (commonly estimated around 66%)
  • Assessment: Question count not published by the exam provider
  • Time limit: 90 minutes
  • Exam fee: $250 (includes one free retake)

Keys to Passing

  • Work through all 100 available questions
  • Review every answer and explanation
  • Track weak areas and revisit them
  • Use our AI tutor for tough concepts

Prometheus Certified Associate Study Tips from Top Performers

1Spend the most time on PromQL, the largest domain at 28%: master rate vs irate vs increase, deriv and delta for gauges, aggregation with by and without, and histogram_quantile on _bucket series.
2Practice on a live Prometheus and Alertmanager stack; scrape node_exporter and blackbox_exporter, then write queries against real counters, gauges, histograms, and summaries.
3Memorize the four metric types and when to use each: counter for monotonic totals, gauge for up-and-down values, histogram for aggregatable bucketed distributions, and summary for client-side quantiles.
4Understand the alert lifecycle: the for clause holds an alert in pending until the condition persists, then it becomes firing and is sent to Alertmanager.
5Know Alertmanager's building blocks cold: routing trees with matchers and continue, group_by, group_wait, group_interval, repeat_interval, inhibition rules, silences, and receivers.
6Be fluent in the data model and exposition format: series identity is metric name plus labels, watch out for cardinality explosions, and recognize _bucket{le}, _sum, and _count series.

Frequently Asked Questions

What are the current exam facts for the PCA?

The Prometheus Certified Associate is an online, proctored, multiple-choice exam from CNCF and the Linux Foundation. It lasts 90 minutes, costs $250 with one free retake, has no prerequisites, includes a 12-month eligibility window, and the resulting certification is valid for two years.

What is the passing score for the PCA exam?

The Linux Foundation does not officially publish the PCA passing score. Community study guides commonly estimate it at roughly 66%, but treat this as an unofficial figure and aim to comfortably exceed it across all five domains.

Which domain carries the most weight on the PCA?

PromQL is the largest domain at 28%, covering data selection, rate and irate, aggregation over time and dimensions, binary operators, and histogram_quantile. Prometheus Fundamentals (20%) is the next largest, so prioritize query writing and core architecture.

Do I need to know Kubernetes to pass the PCA?

Kubernetes is not strictly required, but it helps, because Kubernetes service discovery and relabeling are common in the observability domain. CNCF suggests PCA candidates may already hold KCNA, CKA, or CKAD, though none are prerequisites.

How long is the PCA certification valid?

The PCA certification is valid for two years from the date you pass. After that period you would need to recertify to keep an active credential, in line with the Linux Foundation's certification policy for the PCA.

What is the best way to prepare for the PCA?

Get hands-on with a real Prometheus and Alertmanager setup: scrape exporters, write PromQL with rate and histogram_quantile, and configure alerting rules and routing. Spend the most time on PromQL since it is 28% of the exam, then drill instrumentation and Alertmanager behavior.