Dash0 provides synthetic metrics to help users analyze logs, spans, and related telemetry data dynamically without requiring predefined metric aggregations.

Synthetic metrics are calculated on-the-fly, based on other telemetry ingested by Dash0. Synthetic metrics are not charged on your bill.

Synthetic metrics can query only data that exists in your organization. Spans, logs and metric data points that have expired their retention period cannot be queried via synthetic metrics.

Key Takeaways / Use Cases

Use synthetic metrics to count and alert logs, spans, span events, metric data points and resources dynamically.
Avoid upfront metric definitions, making observability more flexible.
No additional charges on telemetry already sent to Dash0.
Enable advanced query-building for performance monitoring, anomaly detection, and billing analysis.

Retention and Availability

Synthetic metrics are calculated on the fly and rely on retained logs and spans:

Logs, spans and span events have a 30-day retention period.
Metric data points have a 13-month retention period.

Synthetic metrics do not incur extra costs. In Dash0's perspective, you already pay for them by sending the raw telemetry (spans, span events, log records, metric data points) the synthetic metrics are calculated on.

Available Metrics

Dash0 provides the following synthetic metrics, categorized into counters, gauges, and histograms.

Counters (Track Occurrences Over Time)

Note: Dash0 counters behave differently from traditional metrics. They always start from zero regardless of the query timeframe, and they only represent the increase over time. This ensures that the counters provide a clear view of trends and changes without retaining absolute counts.

Metric	Source	Description
`dash0.logs`	Log records	Counts log occurrences over time, supporting filtering and grouping by log and resource attributes.
`dash0.spans`	Spans (traces)	Counts spans over time, supporting filtering and grouping by span and resource attributes.
`dash0.span.events`	Span events (structured data recorded within a span)	Counts the number of span events in traces, supporting filtering and grouping by span event, span and resource attributes.

Gauges (Track State Over Time)

Metric	Source	Description
`dash0.resources`	Telemetry metadata (resources)	Tracks the presence of resources (e.g., hosts, services, pods) sending telemetry. Supports filtering and grouping by resource attributes. Tracks the existence of resources, not a count. A resource disappears from the metric if it stops sending telemetry.
`dash0.metrics.datapoints`	Ingested metric data points	Tracks the count of metric data points arriving in Dash0. Supports filtering and grouping by metric data point and resource attributes.

Histograms (Track Distributions Over Time)

Metric	Source	Description
`dash0.spans.duration`	Span durations (timings)	Represents the distribution of span durations as a histogram, allowing detailed analysis of latency trends. Supports filtering and grouping by span and resource attributes.

Additional Metrics

`dash0.check.*` (Check Rules)

Metric	Type	Description
`dash0.check.enablement_conditions.value`	GAUGE	Encodes whether a check's enablement conditions passed (`1` = passed, `0` = failed). When a check has multiple enablement conditions, this metric emits one data point per condition, with an additional `dash0.enablement_condition.index` attribute (1-based: 1, 2, 3, ...) to identify which condition it relates to.
`dash0.check.evaluation.outcome`	GAUGE	Encodes the outcome of a check's evaluation as a numeric enum. See Enum Values for possible values.
`dash0.check.evaluation.value`	GAUGE	Encodes the value returned by the check's evaluation.
`dash0.check.status`	GAUGE	Represents the check's current state as a numeric enum. See Enum Values for possible values.
`dash0.check.threshold.critical`	GAUGE	Represents the critical threshold for evaluation.
`dash0.check.threshold.degraded`	GAUGE	Represents the degraded threshold for evaluation.

`dash0.issues` (Failed Checks Tracking)

Metric	Type	Description
`dash0.issue.status`	GAUGE	Encodes the issue status as a numeric enum. See Enum Values for possible values.

Enum Values Reference

Several Dash0 metrics use numeric or string enum values to represent states. Understanding these values is essential for querying, filtering, and alerting.

`dash0.check.status` Enum Values

The dash0.check.status metric encodes the current state of a check rule as a numeric gauge value.

Numeric Value	Status Name	Description
`-Inf`	`inactive`	The check is inactive because enablement conditions are not met or the check is disabled.
`0`	`healthy`	The check is passing. No thresholds have been exceeded.
`0.5`	`pending`	The check is within its trigger grace period. The threshold has been exceeded, but not long enough to transition to degraded or critical state.
`1`	`degraded`	The check has exceeded the degraded threshold and the grace period has elapsed.
`2`	`critical`	The check has exceeded the critical threshold and the grace period has elapsed.

Example query to count checks by status:

promql

1234567891011121314
# Count critical checks
count({otel_metric_name="dash0.check.status"} == 2)

# Count degraded checks
count({otel_metric_name="dash0.check.status"} == 1)

# Count pending checks
count({otel_metric_name="dash0.check.status"} == 0.5)

# Count healthy checks
count({otel_metric_name="dash0.check.status"} == 0)

# Count inactive checks
count({otel_metric_name="dash0.check.status"} == -Inf)

`dash0.check.evaluation.outcome` Enum Values

The dash0.check.evaluation.outcome metric encodes the outcome of a check's most recent evaluation. This differs from dash0.check.status as it represents the immediate evaluation result, not the state after grace periods.

Numeric Value	Outcome Name	Description
`-Inf`	`inactive`	The check evaluation did not run because enablement conditions were not met.
`0`	`healthy`	The evaluation result did not exceed any threshold.
`1`	`degraded`	The evaluation result exceeded the degraded threshold but not the critical threshold.
`2`	`critical`	The evaluation result exceeded the critical threshold.

Note

The dash0.check.evaluation.outcome metric does not include a pending state. The pending state only exists in dash0.check.status and represents the grace period between when a threshold is first exceeded and when the check transitions to degraded or critical.

Example queries:

promql

12345678
# Count critical evaluation outcomes
count({otel_metric_name="dash0.check.evaluation.outcome"} == 2)

# Count degraded evaluation outcomes
count({otel_metric_name="dash0.check.evaluation.outcome"} == 1)

# Count healthy evaluations
count({otel_metric_name="dash0.check.evaluation.outcome"} == 0)

`dash0.issue.status` Enum Values

The dash0.issue.status metric encodes the status of issues created when checks fail.

Numeric Value	Status Name	Description
`1.0`	`degraded`	The issue represents one or more checks in degraded state.
`2.0`	`critical`	The issue represents one or more checks in critical state.

Note

Issues only exist for checks that have failed. There are no healthy, pending, or inactive issue statuses.

Example query to count issues by severity:

promql

12345
# Count critical issues
count({otel_metric_name="dash0.issue.status"} == 2)

# Count degraded issues
count({otel_metric_name="dash0.issue.status"} == 1)

`dash0.failed_check.max_status` Attribute Values

The dash0.failed_check.max_status attribute is used for routing notifications based on the highest severity a failed check has reached. This is an attribute on notifications, not a metric.

Value	Description
`degraded`	The check has reached at least degraded status but never critical.
`critical`	The check has reached critical status at some point during its failure cycle.

Use case: Route critical alerts to PagerDuty while sending all alerts to Slack.

1
Notification Channel Filter: dash0.failed_check.max_status = critical

Tip

Once a check reaches critical status, all subsequent notifications (including resolution notifications) will have dash0.failed_check.max_status=critical, even if the check temporarily returns to degraded state. This ensures resolution notifications reach all channels that received alerts.

`dash0.synthetic_check.outcome` Attribute Values

The dash0.synthetic_check.outcome attribute indicates the result of a synthetic check run. It is derived from the assertion outcome of a synthetic check span. Unlike dash0.check.evaluation.outcome, it has no Inactive state because spans only exist for runs that actually executed.

Value	Description
`Healthy`	All critical assertions passed and no degraded assertions failed.
`Degraded`	All critical assertions passed but one or more degraded assertions failed (typically latency thresholds).
`Critical`	One or more critical assertions failed.

Example queries:

promql

123456789101112131415
# Count failed synthetic check runs (critical)
sum(increase({otel_metric_name="dash0.synthetic_check.runs", dash0_synthetic_check_outcome="Critical"}[5m]))

# Count all non-healthy check runs (degraded + critical)
sum(increase({otel_metric_name="dash0.synthetic_check.runs", dash0_synthetic_check_outcome!="Healthy"}[5m]))

# Calculate synthetic check success rate
sum(increase({otel_metric_name="dash0.synthetic_check.runs", dash0_synthetic_check_outcome="Healthy"}[5m]))
/ 
sum(increase({otel_metric_name="dash0.synthetic_check.runs"}[5m]))

# Compare healthy vs degraded vs critical outcomes
sum by (dash0_synthetic_check_outcome) (
  increase({otel_metric_name="dash0.synthetic_check.runs"}[5m])
)

`dash0.web.event.status.code` Attribute Values

The dash0.web.event.status.code attribute indicates the status of web events captured from browser telemetry.

Value	Description
`ERROR`	The web event completed with an error status (e.g., JavaScript error, failed HTTP request).
`UNSET`	The web event completed successfully or no explicit status was recorded.

Example queries:

promql

12345678910
# Count error events (use dash0.web.errors for error-only queries)
sum(increase({otel_metric_name="dash0.web.errors"}[5m]))

# Count successful events
sum(increase({otel_metric_name="dash0.web.events", dash0_web_event_status_code="UNSET"}[5m]))

# Compare error vs successful events side-by-side
sum by (dash0_web_event_status_code) (
  increase({otel_metric_name="dash0.web.events"}[5m])
)

`dash0.trace.origin.type` Attribute Values

The dash0.trace.origin.type attribute identifies the origin type of a trace, allowing you to filter traces by their source.

Value	Description
`SYNTHETIC_CHECK`	The trace originated from a Dash0 synthetic monitoring check. Use this to filter for or exclude synthetic traffic from your analysis.
`WEB`	The trace originated from a browser / RUM (Real User Monitoring) session captured by the Dash0 Web SDK.
`UNKNOWN`	The trace's origin could not be classified (default fallback).

Example queries:

promql

123456789
# Filter out synthetic check traces to analyze only real user traffic
sum by (service_name) (
  rate({otel_metric_name="dash0.spans", dash0_trace_origin_type!="SYNTHETIC_CHECK"}[5m])
)

# Analyze only synthetic check traces
sum by (service_name) (
  rate({otel_metric_name="dash0.spans", dash0_trace_origin_type="SYNTHETIC_CHECK"}[5m])
)

Tip

When analyzing production performance, consider filtering out dash0.trace.origin.type="SYNTHETIC_CHECK" to exclude synthetic monitoring traffic from your metrics. This ensures your dashboards and alerts reflect real user behavior only.

How-To Guide

Querying Synthetic Metrics in Dash0

Navigate to the Dash0 Query Builder.
Select the synthetic metric you want to analyze.
Apply filters and grouping as needed.
Adjust time ranges to explore historical trends.
Execute the query to visualize results.

Examples

Monitor Performance

P99 Duration filtered by service.name=frontend:

Example query:

promql

1
histogram_quantile(0.99, sum by(service_namespace, service_name) (rate({otel_metric_name = "dash0.spans.duration", service_name = "frontend", service_namespace = "opentelemetry-demo", dash0_operation_name != ""}[$__interval]))) * 1000

Troubleshoot Anomalies

Span Rate using dash0.spans filtered by the otel.span.status_code and grouped by service_name:

Example query:

promql

1
sum by (service_name) (rate({otel_metric_name = "dash0.spans", otel_span_status_code = "ERROR"}[$__interval]))

Log Error Rate Check Rule that calculates the rate of logs with severity=ERROR grouped by service.name:

Example query:

promql

1
sum by (service_name) (rate({otel_metric_name = "dash0.logs", otel_log_severity_range = "ERROR"}[1m])) > $__threshold

Grouping Resources

Example query:

promql

1
sum by (k8s_namespace_name)({otel_metric_name = "dash0.resources"})

Using Synthetic Metrics for Observability

Troubleshoot Anomalies: Identify spikes in logs, spans, and errors.
Monitor Performance: Track latency using dash0.spans.duration.
Detect Infrastructure Issues: Use dash0.resources to identify missing components.

Next Steps

Learn more about the Dash0 Query Builder for advanced filtering.
Integrate synthetic metrics into your alerting pipelines.
Explore distributed tracing with Dash0's span analysis tools.

Synthetic Metrics

Key Takeaways / Use Cases

Retention and Availability

Cost

Available Metrics

Counters (Track Occurrences Over Time)

Gauges (Track State Over Time)

Histograms (Track Distributions Over Time)

Additional Metrics

`dash0.check.*` (Check Rules)

`dash0.issues` (Failed Checks Tracking)

Enum Values Reference

`dash0.check.status` Enum Values

`dash0.check.evaluation.outcome` Enum Values

`dash0.issue.status` Enum Values

`dash0.failed_check.max_status` Attribute Values

`dash0.synthetic_check.outcome` Attribute Values

`dash0.web.event.status.code` Attribute Values

`dash0.trace.origin.type` Attribute Values

How-To Guide

Querying Synthetic Metrics in Dash0

Examples

Monitor Performance

Troubleshoot Anomalies

Grouping Resources

Using Synthetic Metrics for Observability

Next Steps

Further Reading