Prometheus Certified Associate (PCA) Exam Questions 2026

Prometheus Certified Associate (PCA) Exam Questions 2026

Prometheus Certified Associate (PCA) Questions 2026 Contains 560+ exam questions to pass the exam in first attempt.

SkillCertPro offers real exam questions for practice for all major IT certifications.

• For a full set of 570 questions. Go to

https://skillcertpro.com/product/prometheus-certified-associate-pca-exam-questions/

• SkillCertPro offers detailed explanations to each question which helps to understand the concepts better.

• It is recommended to score above 85% in SkillCertPro exams before attempting a real exam.

• SkillCertPro updates exam questions every 2 weeks.

• You will get life time access and life time free updates  

• SkillCertPro assures 100% pass guarantee in first attempt.

Below are the free 10 sample questions.

Question 1:

What should labels represent in metric design?

A. Any data you want to track

B. Unique identifiers for each event

C. Aggregatable dimensions with bounded cardinality

D. Timestamps and values


Answer: C

Explanation:

Label Design Principles

Use labels for low-cardinality dimensions (e.g., method: GET/POST/PUT/DELETE)

Avoid high-cardinality labels (e.g., user_id with millions of unique values)

Keep label values bounded (ideally fewer than 1,000 unique values per label)

Use labels for aggregatable dimensions (e.g., instance, service, region)

Enable easy filtering (e.g., env=prod)

Maintain consistency across metrics

Bad Labels (High Cardinality / Unbounded): ❌

user_id, email, ip_address, request_id, timestamp, path

Good Labels (Low Cardinality / Useful for Aggregation): ✅

method, status, service, region, version

Cardinality Calculation

Total Cardinality = metric_count × label1_values × label2_values × ...

Keep the total number of time series manageable

Ideally < 100,000 series per metric to ensure performance and scalability

Question 2:

What is the difference between recording rules and alerting rules?

A. Recording rules create alerts, alerting rules record data

B. They are the same thing

C. Recording rules pre-compute expensive queries; alerting rules trigger alerts

D. Recording rules store data; alerting rules query data

Answer: C

Explanation:

         Recording Rule Example

record: job:api_requests:rate5m
expr: rate(api_requests_total[5m])

• Precomputes the rate metric

• Enables simpler queries like:

sum(job:api_requests:rate5m)

instead of:

sum(rate(api_requests_total[5m]))

Alerting Rule Example

alert: HighErrorRate
expr: job:api_errors:rate5m > 0.05

Naming Convention for Recording Rules

level:metric:operations

Key Notes

• Both recording and alerting rules are defined in rule files

• These files are loaded using the rule_files configuration in prometheus.yml

• Validate rules using:

promtool check rules

Question 3:

What is a span in distributed tracing?

A. A time window for metric aggregation

B. A single unit of work in a distributed system with timing information

C. A network packet between microservices

D. A log entry that spans multiple lines

Answer: B

Explanation:

Distributed Tracing – Key Concept

In distributed tracing, a span is a named, time-bound operation that represents a single unit of work within a system.

• Each span has a start time and duration

• Spans can include tags (metadata) and logs (event details)

• Spans are linked using parent-child relationships to form a complete trace

A trace shows the full journey of a request as it moves through different services in a distributed system.

Key Insight

While Prometheus focuses on metrics, understanding tracing concepts like spans and traces is essential for achieving comprehensive observability across modern distributed architectures.

Question 4:

What does Prometheus use for service discovery in Kubernetes?

A. kubernetes_sd_config

B. A separate service discovery tool

C. kubectl

D. Kubernetes API directly

Answer: A

Explanation:

         Kubernetes Service Discovery Example

scrape_configs:
  - job_name: kubernetes-pods
kubernetes_sd_configs:
  - role: pod
relabel_configs:
  - source_labels: [__meta_kubernetes_pod_annotation_prometheus_io_scrape]
    action: keep
    regex: true

Metadata Labels

• __meta_kubernetes_pod_name

• __meta_kubernetes_namespace

• __meta_kubernetes_pod_label_*

• __meta_kubernetes_pod_ip

Service Discovery Roles

• pod – Each pod is treated as a target

• service – Kubernetes services

• endpoints – Service endpoints

• node – Cluster nodes

• ingress – Ingress resources

Common Pattern

• Use annotations to control scraping:

  • prometheus.io/scrape

  • prometheus.io/port

  • prometheus.io/path

Key Benefits

• Handles dynamic environments automatically

  • Pods are created/terminated → Prometheus updates targets in real time

Requirements

• Proper RBAC permissions are needed for Prometheus to access the Kubernetes API

Question 5:

In Grafana, what is the purpose of dashboard folders?

A. To store dashboard files

B. To backup dashboards

C. To compress dashboards

D. To organize related dashboards logically

Answer: D

Explanation:

Dashboard Organization Strategies

• By Team:

  • Frontend, Backend, SRE, Database teams each have their own folder

• By Layer:

  • Application, Infrastructure, Network, Security

• By Service:

  • API Dashboards, Database Dashboards, Cache Dashboards

• By Environment:

  • Production, Staging, Development

• Mixed Approach:

  • Example: Production/Backend, Production/Frontend, etc.

Benefits

• Easier Navigation: Quickly find relevant dashboards

• Access Control: Manage permissions at the folder level

• Better Discoverability: Helps new team members locate dashboards easily

Best Practices

• Use consistent naming conventions

• Avoid creating too many folders (keep structure simple)

• Use tags alongside folders for better filtering

• Document the purpose of each folder

Provisioning

• Define dashboard folders as code

• Store and manage structure using version control

• For a full set of 570 questions. Go to

https://skillcertpro.com/product/prometheus-certified-associate-pca-exam-questions/

• SkillCertPro offers detailed explanations to each question which helps to understand the concepts better.

• It is recommended to score above 85% in SkillCertPro exams before attempting a real exam.

• SkillCertPro updates exam questions every 2 weeks.

• You will get life time access and life time free updates  

• SkillCertPro assures 100% pass guarantee in first attempt.

Question 6:

What does the timestamp() function return?

A. Time series timestamp metadata

B. Timestamp of each sample in the vector

C. Current server timestamp

D. Time since epoch in seconds

Answer: B

Explanation:

timestamp() Use Cases

• Freshness Check:

time() - timestamp(metric) > 300

→ Detects if a metric is older than 5 minutes

• Scrape Timing:

timestamp(up)

→ Shows the last successful scrape time

• Staleness Detection:

time() - timestamp(metric)

→ Alert when this exceeds a defined threshold

• Scrape Rate Validation:

rate(timestamp(metric)[5m])

→ Should be approximately 1 (one sample per scrape interval)

Comparison with Other Time Functions

• time() → Current query evaluation time

• minute(), hour() → Extract time components

Combining with vector()

timestamp(metric) - vector(time() - 3600)

→ Compares metric timestamp to 1 hour ago

Key Insight

• timestamp() is essential for monitoring data freshness, scrape health, and metadata timing in Prometheus

Question 7:

Which function would you use to calculate the 95th percentile of a histogram metric?

A. topk()

B. avg()

C. quantile()

D. histogram_quantile()

Answer: D

Explanation:

histogram_quantile() Overview

histogram_quantile() is used to calculate quantiles (percentiles) from Prometheus histogram metrics.

Syntax

histogram_quantile(φ, histogram_metric)

• φ ranges from 0 to 1

  • 0.5 → Median (50th percentile)

  • 0.95 → 95th percentile

  • 0.99 → 99th percentile

Correct Usage

• Always apply rate() or increase() first:

histogram_quantile(0.99, rate(latency_bucket[5m]))

Requirements

• Metric must have the _bucket suffix

• Must include the le (less-than-or-equal) label

• The +Inf bucket must be present

Aggregation Pattern

• When aggregating across labels, use:

histogram_quantile(0.95, sum by (le) (rate(latency_bucket[5m])))

Key Insight

• Enables accurate percentile calculations (e.g., latency SLOs) from histogram-based metrics in Prometheus

Question 8:

What is the purpose of the global scrape_interval?

A. To set interval for global alerts

B. To set query interval

C. To set the default scraping frequency for all jobs

D. To set retention interval

Answer: C

Explanation:

         scrape_interval Considerations

• Data Granularity:

  • Shorter intervals → finer resolution

  • Longer intervals → coarser data

• System Load:

  • Shorter intervals → more frequent requests to targets

• Storage Usage:

  • Shorter intervals → more samples → higher disk usage

• Alert Sensitivity:

  • Shorter intervals → faster detection of issues

Typical Values

• 10s – 15s: High-resolution, critical systems

• 30s – 1m: Standard (most common)

• 2m – 5m: Low-priority or high-cardinality metrics

Configuration Example

global:
  scrape_interval: 30s

scrape_configs:
  - job_name: critical
scrape_interval: 10s

Key Considerations

• Can override scrape_interval per job in scrape_configs

• Balance observability detail vs resource cost

Important Insight

• Shorter intervals significantly increase storage needs

  • Example: 10s interval = ~6× more data than 1m

Question 9:

What is the purpose of Alertmanager‘s ‘group_wait‘ configuration?

A. Initial wait before sending first notification for a group

B. Wait time before evaluating alerts

C. Wait between groups

D. Wait for group consensus

Answer: A

Explanation:

Alertmanager Timing Configurations

• group_wait

  • Initial delay before sending the first notification

  • Allows batching of alerts

  • Default: 30s

• group_interval

  • Time to wait before sending updates for an existing alert group

  • Default: 5m

• repeat_interval

  • Time before re-sending a notification for an ongoing alert

  • Default: 4h

Example Scenario

• Situation: 10 instances fail at the same time

• Config:

group_by: [alertname, cluster]
group_wait: 30s

• Behavior:

  • First alert fires

  • Alertmanager waits 30 seconds

  • Sends one notification containing all 10 alerts

• Without grouping:

  • 10 separate notifications ❌

Tuning Considerations

• Shorter group_wait:

  • Faster notifications

  • Less batching → more noise

• Longer group_wait:

  • Better batching

  • Delayed initial notification

Key Insight

• Proper tuning balances urgency (fast alerts) with noise reduction (fewer notifications)

Question 10:

What is the result of: metric1 + metric2 without vector matching modifiers?

A. All metric1 values plus all metric2 values

B. Only series with identical label sets are added

C. All combinations of metric1 and metric2 are added

D. Query error

Answer: B

Explanation:

Vector Matching in Prometheus

Vector matching determines how time series are combined during binary operations.

Matching Types

• Default (no modifiers):

  • All labels must match exactly

• on(labels):

  • Match only on the specified labels

• ignoring(labels):

  • Match on all labels except the specified ones

Examples

• Basic matching:

requests / durations

→ Requires both metrics to have identical labels

• Match only by job:

requests / on(job) durations

Cardinality Handling

• Many-to-one / one-to-many matching:

  • Use group_left or group_right

Important Behavior

• If labels do not match according to the rules → resulting series are dropped

Key Insight

• Vector matching is essential for performing correct and meaningful arithmetic between Prometheus metrics

• For a full set of 570 questions. Go to

https://skillcertpro.com/product/prometheus-certified-associate-pca-exam-questions/

• SkillCertPro offers detailed explanations to each question which helps to understand the concepts better.

• It is recommended to score above 85% in SkillCertPro exams before attempting a real exam.

• SkillCertPro updates exam questions every 2 weeks.

• You will get life time access and life time free updates 

• SkillCertPro assures 100% pass guarantee in first attempt.