Inflight Documentation

Metrics Collector

Multi-APM aggregation hub

The Metrics Collector is the data foundation of the Inflight platform. It aggregates metrics from your existing APM providers, normalizes them into a consistent format, and stores them efficiently across multiple tiers—giving you a unified view of your entire infrastructure regardless of which monitoring tools you use.

Why the Metrics Collector Matters

Most organizations use multiple APM tools across different teams and services. The Metrics Collector unifies this fragmented data into a single source of truth.

4+

Multi-APM Support

Aggregate metrics from Datadog, Dynatrace, New Relic, SigNoz, and more

3-Tier

Intelligent Storage

Hot, warm, and cold storage tiers optimize for both speed and cost

100%

Always Normalized

Consistent metric formats regardless of source APM provider

The Data Fragmentation Problem

When your monitoring data is scattered across multiple tools with different formats, getting a complete picture of your system becomes nearly impossible:

Without Unified Collection

  • Metrics siloed in different APM dashboards
  • Inconsistent naming and units across providers
  • Manual correlation between data sources
  • Limited historical data due to APM retention policies

With the Metrics Collector

  • Single unified view of all metrics
  • Consistent format regardless of source
  • Automatic correlation and enrichment
  • Extended retention with tiered storage

Core Capabilities

The Metrics Collector provides the data infrastructure that powers all of Inflight's optimization and analysis features.

Multi-APM Aggregation

Connect to multiple APM providers simultaneously. Your metrics are unified into a single, consistent view regardless of which monitoring tool collected them.

Datadog integrationDynatrace integrationNew Relic integrationSigNoz integration

Intelligent Normalization

Different APMs report metrics differently. The Collector normalizes everything into a consistent format, making cross-provider analysis seamless.

Consistent metric namingUnit standardizationRuntime-aware mappingConfigurable rules

Anomaly Detection

Automatically identifies unusual patterns in your metrics. Get early warnings about potential issues before they impact your users.

Automatic baseline learningDeviation alertsTrend analysisSeasonality awareness

Gap Handling

Missing data points don't derail your analysis. Intelligent interpolation fills gaps while clearly marking estimated values.

Automatic gap detectionSmart interpolationData quality scoringCompleteness tracking

3-Tier Storage Architecture

Not all data needs the same access speed. The Collector automatically moves metrics through storage tiers based on age and access patterns.

Hot Tier

Recent metrics for real-time dashboards and immediate analysis

Retention:Hours to days
Access:Sub-second queries

Warm Tier

Historical data for trend analysis and model calibration

Retention:Weeks to months
Access:Fast analytical queries

Cold Tier

Long-term archival for compliance and historical reference

Retention:Months to years
Access:Batch retrieval

Data automatically flows from hot to warm to cold based on configurable policies. You control the retention at each tier based on your needs and budget.

Service Profiles

Different services have different metric requirements. Service profiles define which metrics matter for each service type and how they should be collected.

Metric Contracts

Define which metrics are required, optional, or computed for each service. Get alerts when expected metrics are missing.

Quality Scoring

Track data completeness and quality over time. Identify gaps before they impact your optimization recommendations.

Runtime-Aware Collection

JVM services collect GC and heap metrics. Go services collect GOGC and memory limit data. Each runtime gets relevant metrics.

Tier-Based Policies

Critical services get higher collection frequency and longer retention. Less critical services can use more cost-effective policies.

Data Quality Assurance

Optimization recommendations are only as good as the data behind them. The Metrics Collector ensures your data is complete, accurate, and reliable.

Completeness Tracking

Know exactly what percentage of expected data points were received

Anomaly Flagging

Unusual values are flagged for review rather than blindly accepted

Source Attribution

Every data point traces back to its source APM and collection time

Powers the Entire Platform

The Metrics Collector is the foundation that enables all other Inflight components:

Inflight Agent

Registers services and sends configuration data to the Collector

Simulator

Uses collected metrics to calibrate prediction models

AI Advisor

Analyzes normalized metrics to generate optimization recommendations

Monitoring Service

Receives anomaly alerts detected during metric collection

Ready to Unify Your Metrics?

Connect your APM providers and start getting a unified view of your entire infrastructure.