PCA Practice Test 20 — 15 Questions

Question 1

You are monitoring a counter metric 'errors_total' and need to calculate the total number of errors that occurred over the last 1 hour. Which function is most appropriate?

Accepted Answer

increase(errors_total[1h]). The increase() function returns the total increase in the time series over the specified range vector, properly accounting for counter resets.

Answer

rate(errors_total[1h])

Answer

sum(errors_total[1h])

Answer

delta(errors_total[1h])

Question 2

Which TWO of the following should be considered when defining metric names?

Accepted Answer

Use descriptive, clear names. Names should be descriptive and use the unit suffix as per conventions.

Answer

Avoid using underscores

Answer

Use random IDs to ensure uniqueness

Answer

Include the instance name in the metric

Question 3

Which file format is used to define Prometheus alerting rules?

Accepted Answer

YAML. Prometheus alerting rules are defined in YAML format. The rule files are specified in the Prometheus configuration and contain groups of rules with expressions and durations. YAML is the standard for Prometheus configuration files.

Answer

JSON

Answer

TOML

Answer

XML

Question 4

A dashboard shows a latency spike that doesn't correlate with CPU usage. What should you investigate next?

Accepted Answer

Disk I/O and network saturation.. Latency that is decoupled from CPU utilization points to a bottleneck outside of compute processing, most commonly I/O wait or network congestion. Disk I/O saturation (high await, queue depth) and network saturation (packet drops, retransmits, bandwidth limits) can both inflate request latency while CPU sits idle waiting on data. In a Prometheus/Grafana context, you would inspect node_disk_io_time_seconds_total, node_network_transmit_drop_total, or container network metrics to confirm. This is why B is the correct next investigation step.

Answer

Increase the number of replicas.

Answer

Check the memory limit.

Answer

Delete old metrics.

Question 5

Which THREE of the following are valid Alertmanager routing tree parameters?

Accepted Answer

receiver. Routes can filter by 'match', 'match_re', and define a 'receiver'.

Answer

server

Answer

auth_token

Question 6

How do you apply a global notification delay in Alertmanager?

Accepted Answer

By using the group_wait field. The 'group_wait' parameter at the root route level defines the time to wait before sending an initial notification for a new alert group.

Answer

By setting the repeat_interval

Answer

By creating a silence for the first 5 minutes

Answer

By modifying the Prometheus rule interval

Question 7

You are instrumenting a Go application. What is the standard way to register a metric?

Accepted Answer

prometheus.MustRegister(). The prometheus.MustRegister() function is the standard way to register custom metrics with the default registry.

Answer

prometheus.Enable()

Answer

prometheus.Set()

Answer

prometheus.Add()

Question 8

Which TWO of the following are common reasons to use the Pushgateway?

Accepted Answer

For batch jobs that run for seconds. Option A is correct because short-lived batch jobs that run only for seconds may finish before Prometheus can scrape them, so they push their final metrics to the Pushgateway where Prometheus can collect them later. Option E is correct because when the application or its metrics endpoint cannot be reached directly by the Prometheus server—due to network segmentation, firewalls, or NAT—the application can push metrics to a Pushgateway that Prometheus is able to scrape. Option B is wrong because long-running servers are the standard use case for direct Prometheus scraping, not the Pushgateway. Option C is wrong because the Pushgateway is a metrics cache, not a reverse proxy for arbitrary traffic. Option D is wrong because Prometheus and the Pushgateway store metrics in their own time-series format, not in SQL databases.

Answer

To monitor long-running servers

Answer

To act as a reverse proxy

Answer

When metrics need to be stored in SQL

Question 9

What is the benefit of using recording rules?

Accepted Answer

They improve query performance. Recording rules precompute expensive or frequently-used PromQL expressions and save the result as a new time series, so subsequent queries read the precomputed series instead of re-evaluating the full expression over raw data. This dramatically improves query performance, especially for dashboards and alerts that run the same heavy aggregation repeatedly. They are defined in rule files and evaluated at a configurable `interval`.

Answer

They enable alert notifications

Answer

They bypass TSDB

Answer

They reduce storage usage

Question 10

Which TWO of the following label matching operators are supported in Prometheus label selectors? (Choose two)

Accepted Answer

= (exact equality). Prometheus label selectors support the exact equality operator =, as in http_requests_total{job="api"}, which matches series whose label value is exactly the given string, so option A is correct. They also support the regular expression match operator =~, as in http_requests_total{job=~"api|web"}, which matches series whose label value matches the anchored RE2 regular expression, so option B is correct. The remaining operators are not part of PromQL selector syntax: #= is not a wildcard operator, == is a comparison operator used in expressions rather than label matching, and := is not a Prometheus label matching operator at all.

Answer

#= (wildcard match)

Answer

== (boolean equality comparison)

Answer

:= (assignment and match)

Question 11

How does the 'stale marker' work in Prometheus?

Accepted Answer

It prevents the graph from showing data after the target goes down.. When a target stops responding, Prometheus writes a stale marker to the TSDB so that graphs stop at the last known value rather than showing an infinite line or incorrect data.

Answer

It restarts the scrape job.

Answer

It sends an alert.

Answer

It deletes the metric data.

Question 12

What is the primary difference between a metric and a log?

Accepted Answer

Metrics represent snapshots of state; logs represent discrete events.. Metrics are numeric time-series measurements (counters, gauges, histograms) sampled at intervals, so each data point is a snapshot of a value at a moment in time. Logs are timestamped, discrete event records — text or structured payloads emitted when something happens (a request, an error, a state change). This distinction drives everything downstream: metrics are aggregated and cheap to query at scale, while logs preserve per-event detail for forensics.

Answer

Metrics are for debugging; logs are for alerting.

Answer

Logs are cheaper to store.

Answer

Logs are always real-time.

Question 13

A Prometheus alert expression uses rate(http_requests_total[5m]) > 10. During a service restart, counter resets occur. How does the rate() function account for these counter resets?

Accepted Answer

It assumes the counter reset to zero and adds the new value to the previous value.. The rate() function automatically detects decreases in counter values (resets) and adjusts the calculation assuming the counter restarted at zero.

Answer

It treats the negative difference as an error and returns 0 for the entire range.

Answer

It adds a fixed offset value configured in Prometheus global settings.

Answer

It drops the samples surrounding the reset, resulting in a null value for that evaluation.

Question 14

Which THREE of the following modifiers or keywords are used in PromQL binary vector matching? (Choose THREE)

Accepted Answer

group_left. In PromQL binary vector matching, the on keyword (option C) restricts matching to the listed labels, while ignoring (option D) excludes the listed labels from the default matching set; both are valid matching modifiers. The group_left keyword (option A) is a valid many-to-one grouping modifier used with on/ignoring to allow multiple series on the left side to match a single series on the right. The other options are not PromQL keywords: keep_matching (B) does not exist, and group_by (E) is not a binary vector matching modifier (grouping in PromQL uses by/without in aggregations, not group_by).

Answer

keep_matching

Answer

group_by

Question 15

Which golden signal would you monitor to detect if a specific API endpoint is returning 404s?

Accepted Answer

Errors. Errors is the golden signal that measures the rate of requests that fail — including HTTP 5xx, 4xx, and application-level failures. A 404 response from an API endpoint is a client error (4xx) and falls under the Errors signal. Monitoring error rates per endpoint lets you detect misrouted requests, broken links, or misconfigured clients. The four golden signals from Google SRE are Latency, Traffic, Errors, and Saturation.

Answer

Traffic

Answer

Saturation

Answer

Latency