Metrics & Analytics

можно. collects feature flag usage metrics: how many times a flag was evaluated and what result it returned. Metrics are visualized in the web dashboard via sparkline charts and accessible through the REST API.

How Metrics Are Collected

The SDK accumulates evaluation counters for each flag in memory and sends them to the server at a configurable interval (default: 60 seconds):

sequenceDiagram
    participant App
    participant SDK
    participant Server

    loop Flag evaluation
        App->>SDK: isEnabled("new-checkout", ctx)
        SDK-->>App: true/false
        Note over SDK: trueCount++ or falseCount++
    end

    loop Every 60 seconds
        SDK->>Server: POST /api/client/metrics
        Note over Server: Saves to flag_metrics
    end

Each record contains:

Field	Description
flagId	Flag ID
environmentId	Environment ID
evaluationTrueCount	Number of true returns
evaluationFalseCount	Number of false returns
timeBucket	Time interval (rounded to the minute)
instanceId	SDK instance identifier
appName	Application name

Viewing Metrics in the Dashboard

Each flag page displays a sparkline chart — a miniature graph showing the flag's evaluation trend over the selected period. The green line represents true, the gray line false.

Clicking the sparkline opens the Flag Metrics dialog with a full chart and filters:

Filter	Description
Environment	Development / Production
Instance	A specific SDK instance
Application	Application name
Period	Last hour / 24 hours / 7 days / 30 days

Metrics REST API

Flag Metrics

GET /api/v1/flags/{flagId}/metrics

Parameter	Type	Description
environmentId	long	Environment ID
instanceId	string	SDK instance ID
appName	string	Application name

curl "http://localhost:8080/api/v1/flags/42/metrics?environmentId=3" \
  -H "Authorization: Bearer $JWT_TOKEN"

Response:

{
  "metrics": [
    {
      "timeBucket": "2026-06-21T13:41:00Z",
      "trueCount": 150,
      "falseCount": 50,
      "instanceId": "instance-1",
      "appName": "web-app"
    }
  ]
}

Project Metrics

GET /api/v1/metrics

Parameter	Type	Description
environmentId	long	Environment ID

Returns aggregated metrics across all project flags.

Disabling Metrics

If metrics are not needed, disable them in the SDK:

MozhnoConfig config = MozhnoConfig.builder()
    .appName("my-app")
    .instanceId("instance-1")
    .mozhnoUrl("http://localhost:8080")
    .apiKey("<api-key>")
    .disableMetrics(true)
    .build();

Exporting Metrics via SPI

Through MetricsSinkSpi, metrics can be directed to external systems:

Destination	Implementation
Prometheus	Standard endpoint /actuator/prometheus
Datadog, New Relic	Enterprise implementation of MetricsSinkSpi
Local storage	JdbcMetricsSinkProvider (default)

Interpreting Metrics

Pattern	Interpretation
trueCount is growing	The flag is being enabled for more users — rollout is working
falseCount is growing, trueCount = 0	Flag is off or no one matches the targeting rules
Sudden spike in falseCount	Possible issue: flag suddenly stopped being enabled
No metrics at all	SDK is not sending data — check connectivity

Limits

Parameter	Default	Description
Send interval (SDK)	60 seconds	sendMetricsInterval / metricsInterval
Max metrics per API key	1000	CLIENT_MAX_METRICS_PER_KEY

Related Pages

• SDK Overview — how SDK sends metrics
• Monitoring — Prometheus, health checks
• Targeting — rules and segments