← PCA hub

PCA — Alerting & Dashboarding

18% of the PCA exam. Sample questions below; the full library has 20 questions tagged to this domain.

Sample questions on Alerting & Dashboarding

Alerting & Dashboarding

Q1. How do multiple Alertmanager replicas coordinate to deliver notifications exactly once in a highly-available deployment?

Reveal answer and explanations
  1. A A Raft-based leader election picks one instance to send all notifications on behalf of the cluster at any time during peer rotation

    Incorrect. Alertmanager HA is not leader-based; Raft is used by other systems.

  2. B They share a Redis cluster that stores dispatch state, acquiring distributed locks per alert fingerprint before sending notifications

    Incorrect. No external datastore like Redis is required or used.

  3. C Gossip mesh: peers share state, single delivery

    Correct. Alertmanager peers connect via `--cluster.peer` and share state over a gossip protocol, so even though all peers receive alerts from Prometheus, only one sends the notification to each receiver.

  4. D HA is not supported; you must run a single replica and fail over manually via a DNS change during any outage of the active node itself

    Incorrect. HA via clustered gossip is explicitly supported and recommended for production.

Alerting & Dashboarding

Q2. Which of the following recording rule names most closely follows the recommended Prometheus naming convention `level:metric:operations`?

Reveal answer and explanations
  1. A `http_requests_total_rate_5m`

    Incorrect. This flat naming does not follow the `level:metric:operations` convention.

  2. B `rate_5m_http_requests_total_by_job`

    Incorrect. The order and punctuation do not match the colon-separated convention.

  3. C `HTTPRequestsRate5m`

    Incorrect. CamelCase is not part of the documented convention.

  4. D `job:http_requests:rate5m`

    Correct. `job:http_requests:rate5m` encodes the aggregation level (`job`), the metric name (`http_requests`), and the operation (`rate5m`), matching the recommended convention.

Alerting & Dashboarding

Q3. What is 'inhibition' in Alertmanager?

Reveal answer and explanations
  1. A Firing alert suppresses notifications for matching alerts

    Correct. Inhibition rules use source/target matchers so a higher-level alert (e.g. a cluster-wide outage) suppresses more specific alerts (e.g. individual probe failures) while it is firing.

  2. B A rate-limiter that caps outgoing notifications per minute per receiver to protect downstream systems from bursts of pages

    Incorrect. That describes throttling, not inhibition.

  3. C A feature that restricts Prometheus rule evaluation to certain time windows configured globally in the rule manager schedule

    Incorrect. Evaluation scheduling is a Prometheus concern.

  4. D A PromQL function that masks out selected time series from query results based on label matcher expressions at query time

    Incorrect. Inhibition is an Alertmanager configuration, not a PromQL function.

Alerting & Dashboarding

Q4. By default, when a firing alert's condition becomes false and the alert resolves, what does Alertmanager do?

Reveal answer and explanations
  1. A It does nothing; resolution is silent and only firing notifications are ever sent

    Incorrect. Alertmanager supports explicit resolved notifications.

  2. B It only sends resolved notifications when `for: 0s` is configured on the rule

    Incorrect. Resolved sends are orthogonal to the `for` clause.

  3. C Resolved notifications are always suppressed if the alert was ever grouped with another alert

    Incorrect. Grouping does not suppress resolved notifications on its own.

  4. D It sends a resolved notification to each receiver with `send_resolved` enabled (default for several integrations)

    Correct. For receivers with `send_resolved: true` (the default for several integrations), Alertmanager sends a resolved notification when the alert clears, which is essential for on-call tooling state.

Alerting & Dashboarding

Q5. Which of the following notification integrations are available out-of-the-box in Alertmanager?

Reveal answer and explanations
  1. A Only email notifications with SMTP configuration and STARTTLS support for encrypted delivery to SMTP relays configured by the admin

    Incorrect. Many receivers are supported natively.

  2. B Slack, PagerDuty, email, webhook (and more)

    Correct. Alertmanager ships with built-in receiver types for Slack, PagerDuty, email, OpsGenie, VictorOps, Pushover, Telegram, and a generic webhook that can bridge to anything else.

  3. C Only the generic webhook — all other integrations require external plugins or sidecar processes to be deployed to the cluster hosts

    Incorrect. Multiple receiver types are first-class, not plugin-only.

  4. D None — Alertmanager delegates all delivery to Prometheus, which runs the SMTP client and paging integrations itself internally via plugins

    Incorrect. Prometheus forwards alerts to Alertmanager; it does not deliver notifications itself.

Drill Alerting & Dashboarding with the full bankDomain Drill mode targets your weak areas — paid feature

How this domain is tested

Alerting & Dashboarding accounts for 18% of the PCA exam. Expect questions that test recall of terminology and the ability to read short scenarios — not deep configuration. Use the sample questions above as difficulty calibration; if any feel hard, the rest of our 20-question domain bank will close those gaps.