Querying Metrics

SigNoz supports querying metrics using the visual Query Builder in both the Metrics Explorer and Dashboards, with ClickHouse SQL and PromQL available as additional options in Dashboards.

Query Builder

The Query Builder is the primary way to query metrics. It is available in both Metrics Explorer and Dashboards.

Selecting a Metric

Choose the metric you want to query from the Metric dropdown. You can search by name to find the right metric.

Filtering

Use the Filter field to narrow down the data. Filters use metric attributes (labels) to select specific time series. For example:

• service.name = 'frontend' — only data from the frontend service
• k8s.namespace.name IN ('production', 'staging') — data from specific namespaces

Aggregation

Metrics queries use a two-stage aggregation model:

1. Temporal aggregation (Aggregate within time series) — consolidates data points within each time series over time intervals. Available functions depend on the metric type:

   Metric Type	Available Functions
   Gauge	Latest, Sum, Avg, Min, Max, Count, Count Distinct
   Counter	Rate, Increase
   Histogram / Exp. Histogram	P50, P75, P90, P95, P99

2. Spatial aggregation (Aggregate across time series) — merges multiple time series into fewer series. Functions include Sum, Avg, Min, and Max.

For a deeper understanding of metric types and how aggregation works, see Types and Aggregation.

Group By

Use Group By to segment results by one or more attributes. For example, grouping by service.name produces a separate time series for each service.

Having

The Having clause filters results after aggregation. For example, count() > 1000 shows only groups where the count exceeds 1000.

Functions

You can apply additional functions to transform query results:

• Smoothing — EWMA (Exponentially Weighted Moving Average) for noisy data
• Arithmetic — Absolute value, Running Diff, Log2, Log10
• Clamping — Clamp Min/Max to cap values
• Time Shift — Compare current data against a previous time period

Formulas

Plot up to ten queries on the same panel and combine them using formulas. For example, to calculate error rate: (A / B) * 100, where A is error count and B is total count.

For full details on the Query Builder, see the Query Builder documentation.

ClickHouse SQL (Dashboards Only)

When building Dashboard panels, you can write raw ClickHouse SQL queries against the metrics tables. This is useful for advanced analytics that go beyond what the Query Builder supports.

For details on the metrics table schema and example queries, see the ClickHouse Metrics Query guide.

PromQL (Dashboards Only)

Dashboard panels also support PromQL (Prometheus Query Language) for users familiar with Prometheus-style queries. SigNoz supports UTF-8 quoted selector syntax for OpenTelemetry metric names with dots:

{"system.cpu.utilization", "service.name"="frontend"}

For details and examples, see the PromQL guide.