Datadog Data Collection, Resolution, and Retention

Find below a summary of Datadog data collection, resolution, and retention:

Product categorySourceCollection MethodsCollection intervalMinimum ResolutionDefault Retention
APMError TrackingAPM spans (unsampled)2 minutes60 seconds15 days
APMIndexed spans (unsampled)Datadog Agent + tracing library10 seconds1 millisecondPlan
APMProfilesDatadog Agent + tracing library60 seconds60 seconds7 days
APMProfile metricsDatadog Agent + tracing library60 seconds60 seconds1 month
APMServices/resources statistics and span summariesDatadog Agent + tracing library10 seconds10 seconds30 days
APMTraces (sampled)Datadog Agent + tracing library10 seconds1 millisecond15 days. Viewed traces are retained long-term. See Trace Retention for details.
APMTrace metrics (unsampled)Datadog Agent + tracing library10 seconds1 second15 months
ASMSuspicious requestsDatadog Agent + tracing library10 seconds1 millisecond90 days. Underlying traces kept for 15 days.
CI VisibilityPipeline, Stage, Job, Step, Command spanWebhooks, Datadog Agent + pluginData source-dependent1 millisecond15 months.
CD VisibilityDeployment executionWebhooks, Datadog Agent + pluginData source-dependent1 millisecond30 days.
CloudAlibabaAPI crawler10 min (default)1 min15 months
CloudAWSAPI crawler10 min (default)1 min15 months
CloudAzureAPI crawler2 min (default)1 min15 months
CloudGoogle CloudAPI crawler5 min (default)1 min15 months
Cloud Cost ManagementAWSCost and Usage Report1 hour1 day15 months
Cloud Cost ManagementAzureCost Exports1 hour1 day15 months
Cloud Cost ManagementGoogle CloudDetailed Usage Cost Export1 hour1 day15 months
Cloud SIEMSecurity SignalsDatadog Cloud SIEMReal time1 millisecond15 months
Cloud Security ManagementFindingsDatadog Cloud Security Management Misconfigurations15 minutes to 4 hours depending on resource type1 minute15 months
CSM ThreatsSignalsDatadog Cloud Security Management ThreatsReal time1 ms15 months
Database MonitoringQuery MetricsDatadog Agent + enabled integrations10 seconds1 second15 months
Database MonitoringQuery SamplesDatadog Agent + enabled integrations1 minuten/a15 days
DORA MetricsDeployments, FailuresAPI, Datadog productsData source-dependent1 millisecond15 months
DORA MetricsDeployment Frequency, Change Lead Time, Change Failure Rate, Mean time to restoreAPI, Datadog productsData source-dependent1 millisecond15 months
Incident ManagementIncident ManagementIncident datan/an/a∞
InfrastructureAgent integrationsDatadog Agent + enabled integrations15 seconds1 second15 months
InfrastructureCustom metrics (Agent check)Datadog Agent + custom Agent check15 seconds1 second15 months
InfrastructureCustom metrics (API)POST directly to Datadog’s APIReal time1 second15 months
InfrastructureCustom metrics (StatsD)Datadog Agent (built-in statsD collector)15 seconds1 second15 months
InfrastructureEventsDatadog Agent, integrations, or APIReal time1 second15 months
InfrastructureOrchestrator Explorer (Kubernetes)Datadog Agent15 seconds15 seconds15 minutes
InfrastructureLive containersDatadog Agent + enabled Docker integration or Datadog container Agent2 seconds1 second36 hours
InfrastructureLive processesDatadog Agent + Process Agent2 seconds1 second36 hours
InfrastructureNetwork Performance MonitoringSystem Probe30 seconds1 min14 days
InfrastructureSystem metricsDatadog Agent15 seconds1 second15 months
LogsLogsDatadog Agent + Logs, third-party log collectors, or APIReal time1 millisecondPlan
Real User MonitoringError TrackingRUM SDK2 minutes60 seconds30 days
View, Error, and User Action events can be extended to 90 days at no cost by reaching out to Support.
Real User MonitoringReal User MonitoringRUM SDKReal time1 millisecond30 days for Session, View, Action, and Error events
15 days for resource and long task events.
Real User MonitoringSession ReplaysRUM SDKReal time1 millisecond30 days
Service ManagementWorkflow AutomationWorkflow executionsuser configurablen/a30 days for workflow instance input data and output data
Service ManagementService Level ObjectivesDatadog monitors, Datadog Synthetic monitoring, or metrics (infrastructure, APM trace, custom)Data source-dependentData source-dependent3 months for Monitor-based SLOs, matches metric retention duration for Metric-based and Time Slice SLOs (15 months by default)
Software Composition Analysis (SCA)GitHub AppSource coden/an/a15 months
Synthetic MonitoringAPI Test metricsDatadog Synthetic Monitoring applicationUser-defined1 min15 months
Synthetic MonitoringAPI Test resultsDatadog Synthetic Monitoring applicationUser-defined1 minResult seen in the UI by user: 15 months
Result not seen in the UI by user: 2 months
Synthetic MonitoringBrowser Test metricsDatadog Synthetic Monitoring applicationUser-defined5 min15 months
Synthetic MonitoringBrowser Test resultsDatadog Synthetic Monitoring applicationUser-defined5 minResult seen in the UI by user: 15 months
Result not seen in the UI by user: 2 months
Synthetic MonitoringBatchesDatadog Synthetic Monitoring application (through calls to the Synthetics trigger API endpoint or to the Synthetics CI CLI)Depending on calls to the Synthetics trigger API endpoint or to the Synthetic CI CLIn/a15 days
Test OptimizationFlaky testTest Optimization Test spansData source-dependent1 millisecond1 month after the test last flaked
Test OptimizationTest spanDatadog Agent + tracing library60 seconds1 millisecond90 days. Underlying flamegraph kept for 30 days.
USMRED metricsDatadog Agent30 seconds30 second15 months
PREVIEWING: antoine.dussault/service-representation-ga-docs-us1