I85PREP
⌘ K
Datadog Fundamentals study materials
Datadog Fundamentals study resources

Datadog Fundamentals practice questions

One question at a time. Build your Datadog Fundamentals understanding.

First 50 of 199 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Infrastructure Monitoring

Which Datadog Agent command displays the current status of every check running on the host, including errors and execution times?

  1. datadog-agent configcheck
  2. datadog-agent status Correct answer.
  3. datadog-agent diagnose
  4. datadog-agent flare

Explanation

The status subcommand prints a per-check report showing whether each check is running, its last execution time, and any emitted error messages. It is the first place to look when a check is enabled but its metrics are missing.

Question 2Infrastructure Monitoring

Which file holds the main configuration for a host-based Datadog Agent installation on Linux?

  1. /etc/datadog-agent/datadog.yaml Correct answer.
  2. /etc/datadog-agent/datadog.conf
  3. /var/log/datadog/datadog.yaml
  4. /etc/datadog-agent/conf.d/datadog.yaml

Explanation

The Agent's global settings such as api_key, site, tags, and log_level live in datadog.yaml at the root of the configuration directory. Integration-specific files are kept separately under conf.d, and /var/log/datadog holds logs rather than configuration.

Question 3Infrastructure Monitoring

Where do you place the YAML configuration file for an Agent integration check such as the directory check?

  1. /etc/datadog-agent/conf.d Correct answer.
  2. /etc/datadog-agent/checks.d
  3. /etc/datadog-agent/conf.yaml
  4. /opt/datadog-agent/conf.d

Explanation

Each integration's instances and init_config are declared in a YAML file inside conf.d, conventionally named after the integration (for example disk.d/conf.yaml). The Agent reads all of these files at startup and on reload.

Question 4Infrastructure Monitoring

Which transport and port does DogStatsD use by default?

  1. UDP port 8126
  2. TCP port 8126
  3. TCP port 8125
  4. UDP port 8125 Correct answer.

Explanation

DogStatsD listens on UDP port 8125 by default so that custom metrics can be fire-and-forget with minimal overhead. Port 8126 is used by the trace intake for APM, which speaks TCP.

Question 5Infrastructure Monitoring

A check reports the current amount of free memory each time it runs. Which metric submission type should it use?

  1. gauge Correct answer.
  2. rate
  3. histogram
  4. count

Explanation

A gauge is the right submission type for a value that is sampled at a point in time and should be reported as-is, such as free memory or queue depth. The Agent stores the last reported value and does not accumulate or divide it.

Question 6Infrastructure Monitoring

In datadog.yaml, which top-level key holds a list of tags that the Agent applies to all metrics it collects?

  1. hostname
  2. tags Correct answer.
  3. env
  4. log_level

Explanation

The top-level tags key accepts a YAML list of tags that the Agent attaches to everything it submits from that host. This is useful for applying ownership or cost-center tags without editing each integration's configuration.

Question 7Infrastructure Monitoring

Which Agent command packages the current configuration, logs, and status into a file you can send to Datadog support?

  1. datadog-agent flare Correct answer.
  2. datadog-agent restart
  3. datadog-agent info
  4. datadog-agent version

Explanation

The flare command collects logs, resolved check configurations, and status output into an archive intended for support cases. It is the standard way to share a full picture of a host's Agent state without manually gathering several files.

Question 8Infrastructure Monitoring

Which environment variable passes the API key to the Datadog Agent when it runs as a Docker container?

  1. DD_LOGS_ENABLED
  2. DD_SITE
  3. DD_HOSTNAME
  4. DD_API_KEY Correct answer.

Explanation

The containerized Agent reads its API key from the DD_API_KEY environment variable, since editing datadog.yaml inside the image is not practical. Other settings use the DD_ prefix plus the uppercase form of the corresponding YAML key.

Question 9Infrastructure Monitoring

Which Datadog feature lets you inspect individual processes running on a host, with their CPU and memory usage?

  1. Service Map
  2. Live Processes Correct answer.
  3. Host Map
  4. Live Containers

Explanation

Live Processes shows a real-time view of the processes on a host, including per-process CPU and resident memory, which helps spot noisy neighbors. It is fed by data from the Process Agent.

Question 10Infrastructure Monitoring

What is the primary role of Autodiscovery in a containerized environment?

  1. It applies check config templates to containers as they start and stop. Correct answer.
  2. It enables log collection for every container by default.
  3. It scans each container image to determine which integrations are installed inside it.
  4. It discovers new cloud hosts in your AWS account.

Explanation

Autodiscovery watches the container runtime and matches labels or annotations against configuration templates, then starts or stops checks as containers appear and disappear. This keeps monitoring in step with a dynamic, short-lived workload without manual per-container setup.

Question 11Infrastructure Monitoring

In a large Kubernetes cluster, what is a key benefit of running the Datadog Cluster Agent?

  1. It encrypts traffic between node Agents and the Datadog intake.
  2. It replaces node Agents by collecting all host and container metrics centrally.
  3. It serves cluster-level data to node Agents, cutting API server load. Correct answer.
  4. It buffers metrics on disk when the Datadog intake is unreachable.

Explanation

The Cluster Agent queries the Kubernetes API once and distributes cluster-level metadata and check configurations to node Agents, avoiding one API-heavy client per node. It also runs cluster-level checks that only make sense once for the whole cluster.

Question 12Infrastructure Monitoring

What is the retention period for metric data in Datadog?

  1. 1 month (about 30 days)
  2. 6 months
  3. 15 months Correct answer.
  4. 3 months

Explanation

Datadog keeps metrics queryable for 15 months, with resolution becoming coarser over time through automatic rollups. Older data is therefore still graphable, but at a lower granularity than recent data.

Question 13Infrastructure Monitoring

How does Datadog count custom metrics for billing purposes?

  1. Distinct metric name and tag value combinations per hour Correct answer.
  2. The total number of individual data points ingested over the billing month
  3. The number of dashboards that query custom metrics
  4. The number of hosts that submit custom metrics to the account

Explanation

A custom metric is counted as each unique combination of metric name and tag values reported within an hour, and the hourly counts are averaged over the month. Adding high-cardinality tags therefore multiplies the billable metric count.

Question 14Infrastructure Monitoring

A custom metric is tagged with a user ID that has 500,000 unique values. What is the main consequence?

  1. The Agent refuses to send the metric once the number of tag values passes its limit.
  2. The metric is retained for only one hour.
  3. It creates a huge number of distinct time series, raising cost and slowing queries. Correct answer.
  4. Datadog truncates tag values, so the metric loses its tags.

Explanation

Each unique tag combination becomes its own time series, so a high-cardinality tag multiplies storage, custom metric billing, and query cost. Identifiers such as user IDs are better handled through logs, traces, or low-cardinality grouping tags.

Question 15Infrastructure Monitoring

Which three tags make up Datadog's unified service tagging convention?

  1. host, service, env
  2. env, host, region
  3. service, version, team
  4. env, service, version Correct answer.

Explanation

Unified service tagging standardizes env, service, and version as reserved tags so that metrics, traces, and logs can be correlated across the platform. Using this trio consistently is what makes the service-oriented views join data from different products.

Question 16Infrastructure Monitoring

Where does the Agent look for the Python code of a custom check?

  1. /opt/datadog-agent/checks
  2. /var/lib/datadog-agent/checks
  3. /etc/datadog-agent/checks.d Correct answer.
  4. /etc/datadog-agent/conf.d

Explanation

Custom Python checks are placed in checks.d, with the file name matching the check class name, while their instances and init_config go in a matching YAML file under conf.d. The Agent imports the classes from checks.d when it starts.

Question 17Infrastructure Monitoring

You have just edited a check's configuration and want to run that check once immediately to see any errors. Which command do you use?

  1. datadog-agent status <check_name>
  2. datadog-agent reload <check_name>
  3. datadog-agent check <check_name> Correct answer.
  4. datadog-agent configcheck <check_name>

Explanation

The check subcommand executes the named check on demand and prints its output along with any traceback, which is the fastest way to validate a new configuration. The status command only reports the last scheduled run.

Question 18Infrastructure Monitoring

Which page in Datadog lists every metric reporting to your account along with its metadata and tag cardinality?

  1. The Metrics Summary Correct answer.
  2. The Service Map
  3. The Infrastructure List
  4. The Host Map

Explanation

The Metrics Summary is the catalog of all metrics in the account and shows each metric's type, unit, originating integrations, and the number of distinct tag values. It is the standard reference for auditing ingestion and spotting runaway cardinality.

Question 19Infrastructure Monitoring

Which Agent component collects process-level information used by Live Processes?

  1. Security Agent
  2. Trace Agent
  3. Cluster Agent
  4. Process Agent Correct answer.

Explanation

The Process Agent gathers process and container metadata, which powers Live Processes, container views, and process-level metrics. It runs as a separate process from the main collector.

Question 20Infrastructure Monitoring

An Agent registers with a hostname that does not resolve in DNS. Which datadog.yaml key forces a specific hostname?

  1. apm_config.enabled
  2. process_config.enabled
  3. hostname Correct answer.
  4. log_level

Explanation

The hostname key overrides the name the Agent detects and reports, so a host can be pinned to a stable, resolvable name. This is also useful when the auto-detected name would create duplicate hosts.

Question 21Infrastructure Monitoring

A dashboard query spans several weeks, yet the graph returns far fewer points than the raw collection interval would produce. Why?

  1. The Agent only keeps the most recent 24 hours of raw data points in its local cache.
  2. Long-range queries return only the metric's hourly average instead of raw values.
  3. Datadog rolls the data up to a coarser granularity as the time range grows. Correct answer.
  4. Points are dropped whenever a tag has more than 100 distinct values.

Explanation

As the queried window widens, Datadog automatically aggregates stored values into larger buckets so that a manageable number of points is returned. That is why a monthly graph shows minute-level or coarser rollups instead of every raw sample.

Question 22Infrastructure Monitoring

An Agent host must reach the Datadog intake through an outbound HTTP proxy. Which datadog.yaml key is used?

  1. site
  2. proxy Correct answer.
  3. listeners
  4. hostname

Explanation

The proxy key accepts the proxy URL, with optional per-protocol and no-proxy variants for hosts that must be reached directly. The same setting is available to containers through the DD_PROXY_HTTP style environment variables.

Question 23Infrastructure Monitoring

In the Host Map, what do the hexagon size and fill color represent?

  1. Size shows CPU core count; fill shows the cloud region.
  2. Size shows host uptime; fill shows monitor alert status.
  3. Size shows the number of monitors attached to the host; fill shows the owning team.
  4. Size shows the value of a chosen metric; fill shows a chosen tag value. Correct answer.

Explanation

The Host Map groups hosts into hexagons by tags, where the size axis is driven by a metric you select and the fill shade is driven by a tag or metric. This makes it a quick way to spot outliers such as hosts with unusually high or low utilization.

Question 24Infrastructure Monitoring

Two Agents report the same physical machine under different hostnames, which inflates your host count. Which mechanism unifies them?

  1. Metric rollups
  2. Unified service tagging applied to both Agents
  3. Host aliases derived from tags Correct answer.
  4. Autodiscovery templates

Explanation

Host aliases let Datadog recognize that several reported names refer to one machine by linking them through shared tags such as an instance ID or AWS resource ID. Once aliased, the host is counted once and its metrics appear together.

Question 25Infrastructure Monitoring

A custom check sets min_collection_interval to 60 while the Agent's scheduler runs checks every 15 seconds. How often does the check execute?

  1. Once every 75 seconds
  2. Once every 60 seconds Correct answer.
  3. Once every 4 minutes
  4. Once every 15 seconds

Explanation

min_collection_interval sets the shortest interval at which a check may run, so the Agent skips the intervening scheduling cycles until that much time has elapsed. The result is one execution per 60 seconds, not one per scheduler tick.

Question 26Infrastructure Monitoring

An Agent check reads the current number of established TCP connections on a host each time it runs. Which metric type is designed for a value sampled at a single point in time?

  1. count
  2. rate
  3. histogram
  4. gauge Correct answer.

Explanation

A gauge reports the value as it stands at the moment of submission, so it is the correct choice for snapshot measurements that can rise or fall. Datadog keeps the most recent value for the reporting interval rather than summing values across the interval.

Question 27Infrastructure Monitoring

Which Agent command prints the health and state of every check currently running on the host, including recent errors and emitted metric counts?

  1. datadog-agent health
  2. datadog-agent status Correct answer.
  3. datadog-agent configcheck
  4. datadog-agent info

Explanation

The status command reports the running Agent's checks, their execution state, and collected metric counts. It is the standard first place to look when verifying that a check is running correctly.

Question 28Infrastructure Monitoring

Which port does the Agent listen on by default to receive DogStatsD metrics?

  1. 5000
  2. 10514
  3. 8126
  4. 8125 Correct answer.

Explanation

DogStatsD listens on UDP port 8125 by default, and client libraries send custom metrics there. The port can be changed with the dogstatsd_port setting or the corresponding environment variable.

Question 29Infrastructure Monitoring

Which datadog.yaml key selects the Datadog region endpoint, such as datadoghq.eu or us3.datadoghq.com, that the Agent sends data to?

  1. hostname
  2. site Correct answer.
  3. dd_url
  4. api_key

Explanation

The site key chooses the Datadog region and causes the Agent to derive the appropriate intake and API endpoints for that region. It is the normal way to point an Agent at a non-default region.

Question 30Infrastructure Monitoring

A custom check calls self.rate('app.events', 300) once per 60-second run. What value does Datadog store for that interval?

  1. 5, the count divided by the elapsed interval Correct answer.
  2. 0.2, the interval divided by the count
  3. 300, the raw count as submitted
  4. 18,000, the count multiplied by the interval

Explanation

The rate metric type treats the submitted number as a count accumulated since the previous run and converts it to events per second by dividing by the elapsed interval. Here 300 events over 60 seconds becomes 5 per second.

Question 31Infrastructure Monitoring

Which Docker label supplies the list of integration names that Autodiscovery should load for a container?

  1. com.datadoghq.ad.instances
  2. com.datadoghq.ad.logs
  3. com.datadoghq.ad.check_names Correct answer.
  4. com.datadoghq.ad.init_configs

Explanation

The ad.check_names label holds a JSON array naming the checks Autodiscovery should schedule for the container. The matching instances label then supplies the configuration each named check receives.

Question 32Infrastructure Monitoring

A host will be offline for planned maintenance, and you want to silence alerts from its monitors without editing or deleting them. What should you create?

  1. A global mute on each monitor
  2. A threshold override set to a very high value
  3. A deleted monitor with a saved snapshot
  4. A downtime scoped to that host and monitor Correct answer.

Explanation

A downtime suppresses notifications for monitors matching its scope, and scoping it to the host limits the suppression to that host. Monitors remain active and begin evaluating normally after the downtime ends.

Question 33Infrastructure Monitoring

What is the default value, in seconds, of min_collection_interval for an Agent check that does not set it explicitly?

  1. 10
  2. 60
  3. 15 Correct answer.
  4. 300

Explanation

When a check does not specify min_collection_interval, the Agent uses a default of 15 seconds. The setting enforces the minimum spacing between runs of that check, regardless of how often the scheduler wakes up.

Question 34Infrastructure Monitoring

Which set of statuses can a service check submit to Datadog?

  1. pass, fail, skip, and error
  2. up, down, degraded, and flapping
  3. green, yellow, red, and gray
  4. ok, warning, critical, and unknown Correct answer.

Explanation

Service checks report one of four canonical statuses: OK, WARNING, CRITICAL, or UNKNOWN. Monitors built on service checks evaluate these statuses directly.

Question 35Infrastructure Monitoring

Which Agent integration collects JVM heap, thread, and garbage-collection metrics from a Java application?

  1. Windows Performance Counters
  2. expvar
  3. SNMP Correct answer.
  4. JMX

Explanation

The JMX integration connects to a Java application's management interface to collect JVM and application MBeans. It is configured with a list of instances specifying the connection details and the metrics to gather.

Question 36Infrastructure Monitoring

By default, which destination port must be reachable outbound from an Agent host so it can send metrics to Datadog over HTTPS?

  1. 8126
  2. 5000
  3. 443 Correct answer.
  4. 8125

Explanation

The Agent forwards metrics, events, and service checks to Datadog over HTTPS on port 443. Outbound firewall rules must allow this traffic to the intake endpoints for the configured site.

Question 37Infrastructure Monitoring

Which command shows the checks the Agent has loaded together with the configuration each check resolved to, including Autodiscovery templates?

  1. datadog-agent status
  2. datadog-agent configcheck Correct answer.
  3. datadog-agent info
  4. datadog-agent check

Explanation

The configcheck command displays loaded check configurations after templates, environment variables, and Autodiscovery have been resolved. It is useful for confirming that a check is receiving the intended instance settings before debugging collection.

Question 38Infrastructure Monitoring

A custom check measures API latency on 200 hosts, and you need percentile calculations that aggregate across all hosts. Which metric type should it submit?

  1. gauge
  2. distribution Correct answer.
  3. histogram
  4. count

Explanation

Distribution metrics are sent from the Agent without local aggregation, so Datadog can compute percentiles over all reporting hosts and tags. Histogram metrics are aggregated by each Agent locally, so cross-host percentiles are only approximated by combining per-host aggregates.

Question 39Infrastructure Monitoring

Which Agent setting tags DogStatsD metrics with the container and pod they originated from, without the client adding those tags itself?

  1. container_collect_all
  2. use_dogstatsd
  3. dogstatsd_origin_detection Correct answer.
  4. dogstatsd_mapper_profiles

Explanation

Origin detection uses the container identity of the DogStatsD sender to attach container-level tags such as container name and pod to the submitted metrics. Enabling it lets applications send untagged metrics while the Agent enriches them from the container runtime.

Question 40Infrastructure Monitoring

A custom check submits a metric named 'My App.Latency (ms)'. How does Datadog store the metric name?

  1. Exactly as submitted, preserving case and spaces
  2. Lowercased, with spaces replaced by underscores Correct answer.
  3. Rejected, because metric names cannot contain spaces
  4. As 'my_app.latency_(ms)' with the original case preserved

Explanation

Datadog normalizes metric names by converting them to lowercase and replacing spaces with underscores. Matching queries must therefore use the normalized form of the name.

Question 41Application Performance Monitoring

In Datadog APM, what best describes a trace?

  1. A snapshot of host-level CPU and memory usage at a point in time
  2. A single timed operation recorded inside one service during a request
  3. A daily summary of the error rate recorded for a service
  4. A collection of spans that represents one request's end-to-end path Correct answer.

Explanation

A trace is the complete record of one request as it moves through a distributed system, made up of one or more spans. Each span captures a unit of work, and the parent-child relationships between spans reconstruct the request path.

Question 42Application Performance Monitoring

Which three tags make up Datadog unified service tagging?

  1. region, zone, and cluster name
  2. host, container, and pod
  3. env, service, and version Correct answer.
  4. team, owner, and on-call rotation

Explanation

Unified service tagging uses the reserved tags env, service, and version to stitch traces, logs, metrics, and deployments together. The service and env values are typically set per environment while version identifies the deployed build.

Question 43Application Performance Monitoring

By default, which port does the Datadog Agent's trace intake listen on for APM data?

  1. 8127
  2. 8125
  3. 4317
  4. 8126 Correct answer.

Explanation

The Datadog Agent receives traces on TCP port 8126 by default, and tracers are configured to send there. Port 8125 is used for DogStatsD metrics, and 4317 is the OTLP gRPC port.

Question 44Application Performance Monitoring

Which library provides automatic and manual instrumentation for a Python application sending traces to Datadog?

  1. ddapm
  2. dd-trace
  3. datadog-tracer
  4. ddtrace Correct answer.

Explanation

The Python tracing library is published as ddtrace and supports auto-instrumentation plus manual span creation. Other language libraries use different package names, such as dd-trace for Node.js.

Question 45Application Performance Monitoring

How long are traces searchable in APM Live Search before they age out?

  1. 24 hours
  2. 1 hour
  3. 15 minutes Correct answer.
  4. 15 days

Explanation

Live Search streams every incoming trace for a short window of about 15 minutes so you can inspect unsampled traffic. Anything you need to keep beyond that window must be captured by a retention filter and indexed.

Question 46Application Performance Monitoring

How long are spans retained in Datadog after they are captured by a retention filter?

  1. 24 hours
  2. 15 days Correct answer.
  3. 7 days
  4. 30 days

Explanation

Spans that pass a retention filter are indexed and kept for 15 days, during which they can be queried in the Trace Explorer. Live Search, by contrast, only covers roughly the last 15 minutes of traffic.

Question 47Application Performance Monitoring

An engineer wants only failed requests to be indexed for a service. Which retention filter configuration achieves this?

  1. Set the sampling rate to 100% on the default filter
  2. Set the filter query to http.status_code:200
  3. Disable every retention filter so only failures survive
  4. Set the filter query to status:error and enable the filter Correct answer.

Explanation

Retention filters use the trace search query syntax, so a query such as status:error keeps only spans tagged as errors. Enabling that filter while other broad filters are disabled limits indexing to failing traffic.

Question 48Application Performance Monitoring

Where is the head-based sampling decision for a trace made?

  1. In the Datadog backend after full traces are received
  2. In the tracer at the root of the trace, before spans are sent Correct answer.
  3. In the browser SDK only, before any backend call
  4. In the Datadog Agent when spans leave the host

Explanation

Head-based sampling decides at the start of a trace, in the tracing library at the root span, so the decision is propagated to downstream services. This keeps the trace consistent across services and reduces the volume sent to the Agent.

Question 49Application Performance Monitoring

Why do Datadog tracing libraries support 128-bit trace IDs?

  1. To encrypt trace payloads while they are in transit
  2. To allow more than 64 services per organization
  3. To reduce the storage cost of indexed spans
  4. To remain compatible with W3C Trace Context and other tracing systems Correct answer.

Explanation

Many tracing standards and vendors use 128-bit trace identifiers, and W3C Trace Context specifically carries a 32-character hexadecimal trace ID. Emitting 128-bit IDs lets Datadog correlate traces across mixed ecosystems.

Question 50Application Performance Monitoring

Which HTTP headers does Datadog natively use to propagate trace context between services?

  1. x-datadog-trace-id and x-datadog-parent-id Correct answer.
  2. traceparent and tracestate
  3. X-B3-TraceId and X-B3-SpanId
  4. x-request-id and x-correlation-id

Explanation

Datadog's native propagation format injects x-datadog-trace-id, x-datadog-parent-id, and x-datadog-sampling-priority headers. Libraries can also be configured to read and emit W3C or B3 headers for interoperability.