Configure Monitors

Overview

Once you’ve chosen a monitor type, start configuring the monitor. 4 main steps must be completed before saving the monitor :

  • Define the search query: Construct a query to count events, measure metrics, group by one or several dimensions, etc.
  • Set alert conditions: Define alert and warning thresholds , evaluation time frames, and configure advanced alert options.
  • Say what’s happening: Write a custom notification title and message with variables.
  • Notify your team: Choose how notifications are sent to your teams (email, Slack, PagerDuty, etc)

Define the search query

To learn how to construct the search query, see the individual monitor types pages. As you define the search query, the preview graph above the search fields updates.

Alert grouping

Alerts are grouped automatically based on your selection of the group by step when defining your query. If no group is specified, grouping defaults to Simple Alert. If the query is grouped by any dimension, grouping defaults to Multi Alert.

Simple Alert mode aggregate over all reporting sources. You receive one alert when the aggregated value meets the set conditions.

Multi Alert mode apply the alert to each source according to your group parameters. You receive an alert for each group that meets the set conditions. For example, you could group a query looking at a capacity metric by host and device to receive a separate alert for each host device that is running out of space. Note that if your metric is only reporting by host with no device tag, it would not be detected by a monitor group by both host and device. Tag Variables are available for every group evaluated in the multi alert to dynamically fill in notifications with useful context.

Group bySimple alert modeMulti alert mode
(everything)One single group triggering one notificationN/A
1 or more dimensionsOne notification if one or more groups meet the alert conditionsOne notification per group meeting the alert conditions

Set alert conditions

The alert conditions vary based on the monitor type. Configure monitors to trigger if the query value crosses a threshold, or if a certain number of consecutive checks failed.

  • Trigger when the average, max, min, or sum of the metric is
  • above, above or equal to, below, or below or equal to the threshold
  • during the last 5 minutes, 15 minutes, 1 hour, etc. or custom to set a value between 1 minute and 48 hours (1 month for metric monitors)

Aggregation method

The query returns a series of points, but a single value is needed to compare to the threshold. The monitor must reduce the data in the evaluation window to a single value.

OptionDescription
averageThe series is averaged to produce a single value that is checked against the threshold. It adds the avg() function to your monitor query.
maxIf any single value in the generated series crosses the threshold, then an alert is triggered. It adds the max() function to your monitor query.
minIf all points in the evaluation window for your query cross the threshold, then an alert is triggered. It adds the min() function to your monitor query.
sumIf the summation of every point in the series crosses the threshold, then an alert is triggered. It adds the sum() function to your monitor query.

Note: There are different behaviors when utilizing as_count(). See as_count() in Monitor Evaluations for details.

Evaluation window

A monitor can be evaluated using cumulative time windows or rolling time windows. Cumulative time windows are best suited for questions that require historical context, such as “What’s the sum of all the data available up to this point in time?” Rolling time windows are best suited for answering questions that do not require this context, such as “What’s the average of the last N data points?”

The figure below illustrates the difference between cumulative and rolling time windows.

Two graphs showing cumulative vs. rolling time windows. Cumulative time windows continue to expand as time goes on. Rolling time windows cover particular moments in time.

Rolling time windows

A rolling time window has a fixed size and moves its starting point over time. Monitors support looking back at the last 5 minutes, 15 minutes, 1 hour, or over a custom specified time window.

Cumulative time windows

A cumulative time window has a fixed starting point and expands over time. Monitors support three different cumulative time windows:

  • Current hour: A time window with a maximum of one hour starting at a configurable minute of an hour. For example, monitor amount of calls an HTTP endpoint receives in one hour starting at minute 0.
  • Current day: A time window with a maximum of 24 hours starting at a configurable hour and minute of a day. For example, monitor a daily log index quota by using the current day time window and letting it start at 2:00pm UTC.
  • Current month: Looks back at the current month starting on the first of the month at midnight UTC. This option represents a month-to-date time window and is only available for metric monitors.
Screenshot of how a cumulative window is configured in the Datadog interface. The user has searched for aws.sqs.number_of_messages_received. The options are set to evaluate the SUM of the query over the CURRENT MONTH.

A cumulative time window is reset once its maximum time span is reached. For example, a cumulative time window looking at the current month resets itself on the first of each month at midnight UTC. Alternatively, a cumulative time window of current hour, which starts at minute 30, resets itself every hour. For example, at 6:30am, 7:30am, 8:30am, etc.

Evaluation frequency

The evaluation frequency defines how often Datadog performs the monitor query. For most configurations, the evaluation frequency is 1 minute, which means that every minute, the monitor queries the selected data over the selected evaluation window and compares the aggregated value against the defined thresholds.

Evaluation frequencies depend on the evaluation window that is being used. A longer window results in lower evaluation frequencies. The table below illustrates how the evaluation frequency is controlled by larger time windows:

Evaluation Window RangesEvaluation Frequency
window < 24 hours1 minute
24 hours <= window < 48 hours10 minutes
window >= 48 hours30 minutes

Thresholds

Use thresholds to set a numeric value for triggering an alert. Depending on your chosen metric, the editor displays the unit used (byte, kibibyte, gibibyte, etc).

Datadog has two types of notifications (alert and warning). Monitors recover automatically based on the alert or warning threshold but additional conditions can be specified. For additional information on recovery thresholds, see What are recovery thresholds?. For example, if a monitor alerts when the metric is above 3 and recovery thresholds are not specified, the monitor recovers once the metric value goes back below 3.

OptionDescription
Alert threshold (required)The value used to trigger an alert notification.
Warning thresholdThe value used to trigger a warning notification.
Alert recovery thresholdAn optional threshold to indicate an additional condition for alert recovery.
Warning recovery thresholdAn optional threshold to indicate an additional condition for warning recovery.

As you change a threshold, the preview graph in the editor displays a marker showing the cutoff point.

Thresholds preview graph

Note: When entering decimal values for thresholds, if your value is <1, add a leading 0 to the number. For example, use 0.5, not .5.

A check alert tracks consecutive statuses submitted per check grouping and compares it to your thresholds. Set up the check alert to:

  1. Trigger the alert after selected consecutive failures: <NUMBER>

    Each check run submits a single status of OK, WARN, or CRITICAL. Choose how many consecutive runs with the WARN and CRITICAL status trigger a notification. For example, your process might have a single blip where connection fails. If you set this value to > 1, the blip is ignored but a problem with more than one consecutive failure triggers a notification.

    Check thresholds Alert/Warn
  2. Resolve the alert after selected consecutive successes: <NUMBER>

    Choose how many consecutive runs with the OK status resolves the alert.

    Check thresholds Recovery

See the documentation for process check, integration check, and custom check monitors for more information on configuring check alerts.

Advanced alert conditions

No data

Notifications for missing data are useful if you expect a metric to always be reporting data under normal circumstances. For example, if a host with the Agent must be up continuously, you can expect the system.cpu.idle metric to always report data.

In this case, you should enable notifications for missing data. The sections below explain how to accomplish this with each option.

There are two ways to deal with missing data:

  • Metric-based monitors using the limited Notify no data option
  • The On missing data option is supported by APM Trace Analytics, Audit Logs, CI Pipelines, Error Tracking, Events, Logs, and RUM monitors

Do not notify if data is missing or Notify if data is missing for more than N minutes.

You are notified if data is missing or if data is not missing. The notification occurs when no data was received during the configured time window.

Note: It is recommended that you set the missing data window to at least two times the evaluation period.

If you are monitoring a metric over an auto-scaling group of hosts that stops and starts automatically, notifying for no data produces a lot of notifications.

In this case, you should not enable notifications for missing data. This option does not work if it is enabled at a time when data has not been reporting for a long period.

Simple Alert

For a monitor that does not notify on missing data, the monitor skips evaluations and stays green until data returns that would change the status from OK.

Multi Alert

For a monitor that does not notify on missing data, if a group does not report data, the monitor skips evaluations and eventually drops the group. During this period, the bar in the results page stays green. When there is data and groups start reporting again, the green bar shows an OK status and backfills to make it look like there was no interruption.

If data is missing for N minutes, select an option from the dropdown menu:

No Data Options
  • Evaluate as zero / Show last known status
  • Show NO DATA
  • Show NO DATA and notify
  • Show OK.

The selected behavior is applied when a monitor’s query does not return any data. Contrary to the Do not notify option, the missing data window is not configurable.

OptionMonitor status & notification
Evaluate as zeroEmpty result is replaced with zero and is compared to the alert/warning thresholds. For example, if the alert threshold is set to > 10, a zero would not trigger that condition, and the monitor status is set to OK.
Show last known statusThe last known status of the group or monitor is set.
Show NO DATAMonitor status is set to NO DATA.
Show NO DATA and notifyMonitor status is set to NO DATA and a notification is sent out.
Show OKMonitor is resolved and status is set to OK.

The Evaluate as zero and Show last known status options are displayed based on the query type:

  • Evaluate as zero: This option is available for monitors using Count queries.
  • Show last known status: This option is available for monitors using any other query type than Count, for example Gauge, Rate, and Distribution.

Auto resolve

[Never], After 1 hour, After 2 hours, etc. automatically resolve this event from a triggered state.

Auto-resolve works when data is no longer being submitted. Monitors do not auto-resolve from an ALERT or WARN state if data is still reporting. If data is still being submitted, the renotify feature can be utilized to let your team know when an issue is not resolved.

For some metrics that report periodically, it may make sense for triggered alerts to auto-resolve after a certain time period. For example, if you have a counter that reports only when an error is logged, the alert never resolves because the metric never reports 0 as the number of errors. In this case, set your alert to resolve after a certain time of inactivity on the metric. Note: If a monitor auto-resolves and the value of the query does not meet the recovery threshold at the next evaluation, the monitor triggers an alert again.

In most cases this setting is not useful because you only want an alert to resolve once it is actually fixed. So, in general, it makes sense to leave this as [Never] so alerts only resolve when the metric is above or below the set threshold.

Group retention time

You can drop the group from the monitor status after N hours of missing data. The maximum length of time is 72 hours.

Group Retention Time Option

Similar to the Auto-resolve option, the group retention works when data is no longer being submitted. This option controls how long the group is kept in the monitor’s status once data stops reporting. By default, a group keeps the status for 24 hours before it is dropped. The start time of the group retention and the Auto-resolve option are identical as soon as the monitor query returns no data.

Some use cases to define a group retention time include:

  • When you would like to drop the group immediately or shortly after data stops reporting
  • When you would like to keep the group in the status for as long as you usually take for troubleshooting

Note: The group retention time option requires a multi alert monitor that supports the On missing data option. These monitor types are APM Trace Analytics, Audit Logs, CI Pipelines, Error Tracking, Events, Logs, and RUM monitors.

New group delay

Delay the evaluation start by N seconds for new groups.

The time (in seconds) to wait before starting alerting, to allow newly created groups to boot and applications to fully start. This should be a non-negative integer.

For example, if you are using containerized architecture, setting a group delay prevents monitor groups scoped on containers from triggering due to high resource usage or high latency when a new container is created. The delay is applied to every new group (which has not been seen in the last 24 hours) and defaults to 60 seconds.

The option is available with multi alert mode.

Evaluation delay

Delay evaluation by N seconds.

The time (in seconds) to delay evaluation. This should be a non-negative integer. So, if the delay is set to 900 seconds (15 minutes), the monitor evaluation is during the last 5 minutes, and the time is 7:00, the monitor evaluates data from 6:40 to 6:45. The maximum configurable evaluation delay is 86400 seconds (24 hours).

Note: A 15 minute delay is recommended for cloud metrics which are backfilled by service providers. Additionally, when using a division formula, a 60 second delay is helpful to ensure your monitor evaluates on complete values.