APM metric monitors work like regular metric monitors, but with controls tailored specifically to APM. Use these monitors to receive alerts at the service level on hits, errors, and a variety of latency measures.
Analytics monitors allow you to visualize APM data over time and set up alerts based on Indexed Spans. For example, use an Analytics monitor to receive alerts on a spike in slow requests.
To create an APM monitor in Datadog, use the main navigation: Monitors –> New Monitor –> APM.
Choose between an APM Metrics or Analytics monitor:
An alert is triggered whenever a metric crosses a threshold.
Alert when Requests per second, Errors per second, Apdex, Error rate, Avg latency, p50 latency, p75 latency, p90 latency, or p99 latency
is above, above or equal to, below, or below or equal to
Alert threshold <NUMBER>
Warning threshold <NUMBER>
over the last 5 minutes, 15 minutes, 1 hour, etc. or custom to set a value between 1 minute and 48 hours.
An alert is triggered whenever a metric deviates from an expected pattern.
For Requests per second, Errors per second, Apdex, Error rate, Avg latency, p50 latency, p75 latency, p90 latency, or p99 latency
Alert when <ALERT_THRESHOLD>%, <WARNING_THRESHOLD>%
of values are <NUMBER> deviations above or below, above, or below
the prediction during the past 5 minutes, 15 minutes, 1 hour, etc. or custom to set a value between 1 minute and 48 hours.
Advanced alert conditions
For detailed instructions on the advanced alert options (no data, evaluation delay, etc.), see the Monitor configuration page. For the metric-specific option full data window, see the Metric monitor page.
Note: There is a default limit of 1000 Trace Analytics monitors per account. Contact Support to lift this limit for your account.
Define the search query
Construct a search query using the same logic as a trace search.
Monitor over a trace count: Use the search bar (optional) and do not select a facet or measure. Datadog evaluates the number of traces over a selected time frame and then compares it to the threshold conditions.
Monitor over a facet or measure: If a facet is selected, the monitor alerts over the Unique value count of the facet. If a measure is selected, then it’s similar to a metric monitor, and aggregation needs to be selected (min, avg, sum, median, pc75, pc90, pc95, pc98, pc99, or max).
Group traces by multiple dimensions (optional):
All traces matching the query are aggregated into groups based on the value of up to four facets.
Configure the alerting grouping strategy (optional):
Simple alert: Simple alerts aggregate over all reporting sources. You receive one alert when the aggregated value meets the set conditions.
If the query has a group by and you select simple alert mode, you get one alert when one or multiple groups' values breach the threshold. This strategy may be selected to reduce notification noise.
Multi alert: Multi alerts apply the alert to each source according to your group parameters. An alerting event is generated for each group that meets the set conditions. For example, you could group a query by @resource.name to receive a separate alert for each resource when a span’s error rate is high.
Note: Analytics monitors can only be created based on Indexed Spans.
Select alert conditions
Trigger when the metric is above or above or equal to
the threshold during the last 5 minutes, 15 minutes, 1 hour, etc. or custom to set a value between 5 minutes and 48 hours.
Alert threshold: <NUMBER>
Warning threshold: <NUMBER>
Advanced alert conditions
For detailed instructions on the advanced alert options (evaluation delay, etc.), see the Monitor configuration page.
For detailed instructions on the Say what’s happening and Notify your team sections, see the Notifications page.
Note: Find service level monitors on the Services page and on the Service Map, and find resource level monitors on the individual resource pages (you can get there by clicking on the specific resource listed on the Services page).
Additional helpful documentation, links, and articles: