PCA Practice Test 12 — 15 Questions

Question 1

Which THREE of the following are valid set operators in PromQL for combining instant vectors? (Choose THREE)

Accepted Answer

unless. In PromQL, the set operators available for combining instant vectors are `and`, `or`, and `unless`. Option D (`and`) is correct because it returns the elements of the left-hand vector that also exist in the right-hand vector, matching on labels. Option E (`or`) is correct because it returns all elements of the left-hand vector plus any elements from the right-hand vector that do not have matching label sets on the left. Option C (`unless`) is correct because it returns elements of the left-hand vector that do not have matching label sets in the right-hand vector. Options A (`intersect`) and B (`union`) are not valid PromQL set operators; the intersection behavior is expressed by `and`, and union-like behavior is expressed by `or`, so neither keyword exists in the language.

Answer

intersect

Answer

union

Question 2

Which TWO of the following are true about the node_exporter?

Accepted Answer

It runs on port 9100 by default. Option C is correct because node_exporter's default listen address is :9100, so Prometheus scrapes host metrics from http://<host>:9100/metrics unless overridden with --web.listen-address. Option E is correct because node_exporter's collectors read from /proc and /sys on Linux to expose kernel-level metrics such as node_cpu_seconds_total, node_memory_MemAvailable_bytes, node_load1, and node_disk_io_time_seconds_total. Option A is wrong because node_exporter is a host/OS metrics exporter, not a database exporter; database monitoring uses exporters like mysqld_exporter, postgres_exporter, or oracledb_exporter. Option B is wrong because node_exporter is a passive exporter that only responds when Prometheus scrapes its /metrics endpoint; active probing is done by blackbox_exporter. Option D is wrong because node_exporter is designed for Unix-like systems (Linux, BSD, macOS) and does not support Windows out of the box; Windows host metrics require windows_exporter.

Answer

It is for database monitoring

Answer

It performs active monitoring

Answer

It supports Windows out of the box

Question 3

Which THREE of the following are valid Alertmanager grouping parameters?

Accepted Answer

group_by. In Alertmanager's configuration, the route block supports the grouping parameters group_by, group_wait, and group_interval, which is why options B, C, and D are correct. group_by (B) specifies the labels used to aggregate related alerts into a single notification, such as grouping by 'alertname' or 'cluster'. group_wait (D) sets how long Alertmanager waits to buffer alerts for the same group before sending the initial notification, defaulting to 30s. group_interval (C) defines how long to wait before sending a new notification about a group after the first one, defaulting to 5m. The unmarked options do not belong: alert_timeout (A) is not an Alertmanager parameter (timeouts appear in receivers like webhook or in Prometheus scrape configs), and resolve_delay (E) is not a valid Alertmanager grouping setting either.

Answer

alert_timeout

Answer

resolve_delay

Question 4

What is the behavior of the clamp_max() function when applied to an instant vector?

Accepted Answer

It restricts all values in the vector so that none exceed the specified maximum.. clamp_max(v, max) returns the input vector with every sample value capped at the specified maximum: values above max are replaced by max, while values at or below max pass through unchanged. It does not drop or flag series — it mutates the values in place. This is useful for capping spikes in graphs or enforcing upper bounds in derived metrics.

Answer

It drops any time series whose value exceeds the specified maximum.

Answer

It generates an alert if any time series exceeds the maximum.

Answer

It sets values exceeding the maximum to NaN.

Question 5

What is the default port that Prometheus listens on?

Accepted Answer

9090. 9090 is the well-known default port for the Prometheus server web interface.

Answer

9100

Answer

9093

Answer

8080

Question 6

When implementing a custom exporter, what format must the output follow to be correctly scraped by Prometheus?

Accepted Answer

Prometheus Text Format. Prometheus requires a text-based format where each metric is represented by lines, including name, labels, and values.

Answer

JSON

Answer

XML

Answer

Binary Protobuf

Question 7

Which THREE of the following are valid metric types in Prometheus?

Accepted Answer

Gauge. Prometheus defines exactly four core metric types, and three of them appear here. Option A, Gauge, is correct because a gauge represents a value that can arbitrarily go up or down, such as temperature or current memory usage. Option C, Counter, is correct because a counter is a cumulative metric that only increases (or resets to zero on restart), typically used for counts like total requests served. Option D, Histogram, is correct because a histogram samples observations and counts them into configurable buckets, exposing _bucket, _sum, and _count series for distributions and quantile estimation. Option B, Set, is not a Prometheus metric type; it is not part of the client library metric primitives (the fourth valid type is Summary, which is absent here). Option E, Queue, is also not a metric type; it is merely a data structure concept and has no corresponding Prometheus instrumentation primitive.

Answer

Set

Answer

Queue

Question 8

Which metric type would best capture the distribution of request latencies?

Accepted Answer

Histogram. A Histogram is the correct metric type for capturing the distribution of request latencies because it buckets observations into configurable ranges and exposes cumulative counters per bucket, allowing calculation of percentiles (e.g., p50, p95, p99) via histogram_quantile() in PromQL. Latency is inherently a distribution, not a single value, so bucketing is required to observe spread and tail behaviour. Histograms are also aggregatable across instances, which is essential for fleet-wide latency SLOs.

Answer

Summary

Answer

Gauge

Answer

Counter

Question 9

Which THREE of these represent the Four Golden Signals?

Accepted Answer

Errors. The Four Golden Signals, as defined in Google's SRE guidance, are Latency, Traffic, Errors, and Saturation, so option E (Latency) is correct because it measures the time taken to serve requests, typically distinguishing successful-request latency from failed-request latency. Option C (Traffic) is correct because it quantifies demand on the system, such as HTTP requests per second or network I/O rate. Option B (Errors) is correct because it captures the rate of failed requests, including explicit failures like HTTP 5xx responses and implicit failures such as wrong content returned with a 200 status. The remaining options do not belong: Memory Usage (A) is only one possible component of saturation rather than a golden signal itself, and Availability (D) is a broader service-level objective concept, not one of the four signals.

Answer

Memory Usage

Answer

Availability

Question 10

What is the purpose of a 'grouping' configuration in the Alertmanager 'route' block?

Accepted Answer

To aggregate similar alerts into a single notification. Grouping categorizes multiple alerts into a single notification based on common label sets, reducing alert fatigue.

Answer

To inhibit alerts based on severity

Answer

To ensure alerts are processed in alphabetical order

Answer

To create hierarchical alert routing

Question 11

Which THREE built-in functions in PromQL are designed specifically to analyze changes over time for gauge metrics? (Choose three)

Accepted Answer

delta(). delta() is correct because it computes the difference between the first and last sample values of a gauge over a specified range, directly measuring change over time. deriv() is correct because it calculates the per-second derivative of a gauge using simple linear regression, quantifying the rate of change of the gauge's value. predict_linear() is correct because it applies linear regression over a gauge's range to extrapolate future values, which is inherently a time-based change analysis for gauges. rate() and increase() are not correct here because they are designed for counters, not gauges — rate() computes the per-second average rate of increase of a counter, and increase() computes the total counter increase over a range, both of which assume monotonically increasing counter semantics.

Answer

rate()

Answer

increase()

Question 12

When using the Prometheus Go client, what is the difference between a Gauge and a GaugeVec?

Accepted Answer

GaugeVec supports labels. In the Prometheus Go client, a Gauge is a metric that represents a single numerical value that can go up or down, while a GaugeVec is a vector of Gauges that supports labels. The GaugeVec allows you to create multiple Gauge metrics differentiated by label values (e.g., by instance, method, or status), enabling multi-dimensional metrics. This is the key difference: GaugeVec supports labels, whereas a plain Gauge does not.

Answer

Gauge is faster

Answer

GaugeVec is for Histograms

Answer

There is no difference

Question 13

When using the 'repeat_interval' in Alertmanager, what does it define?

Accepted Answer

How often to resend firing alerts. The repeat_interval defines how long to wait before re-sending a notification for an alert that is still active.

Answer

The timeout for the receiver

Answer

The TTL of the alert

Answer

How long to wait before firing

Question 14

What does the `_count` suffix denote in a summary or histogram?

Accepted Answer

The total number of samples. The _count suffix represents the total number of observations made for that specific metric.

Answer

The maximum value

Answer

The sum of values

Answer

The current bucket value

Question 15

When running node_exporter, how can you disable specific collectors like 'textfile'?

Accepted Answer

--no-collector.textfile. Collectors can be disabled using the --no-collector.<name> flag.

Answer

collector.textfile=false

Answer

exclude=textfile

Answer

--disable-textfile