OpenTelemetry

New here? Start with the Quickstart to launch Icelake + Grafana locally and send your first telemetry in 5 minutes.

Overview

Icelake speaks OTLP protobuf natively for logs, metrics, and traces. Ship observability data from any OTel-instrumented application or from the OpenTelemetry Collector — no adapters in between.

All OTLP endpoints authenticate via HTTP Basic auth — username is your client ID, password is your API key (ilk_ + 64 hex characters).

Getting started

1. Configure the OpenTelemetry Collector

Point your OTel Collector’s otlphttp exporter at Icelake:

otel-collector-config.yaml

exporters:
  otlphttp:
    endpoint: "https://api.icelake.eu"
    auth:
      authenticator: basicauth/icelake

extensions:
  basicauth/icelake:
    client_auth:
      username: "your-client-id"
      password: "ilk_your-api-key-here"

service:
  extensions: [basicauth/icelake]
  pipelines:
    logs:
      receivers: [otlp]
      processors: [batch]
      exporters: [otlphttp]
    metrics:
      receivers: [otlp]
      processors: [batch]
      exporters: [otlphttp]
    traces:
      receivers: [otlp]
      processors: [batch]
      exporters: [otlphttp]

2. Verify ingestion

Check that data is flowing:

# Health check
curl https://api.icelake.eu/health

# Query ingested OTel logs via the Loki-compatible API
curl -u "your-client-id:ilk_your-api-key-here" \
  "https://api.icelake.eu/loki/api/v1/query?query={source=\"otel\"}"

OTLP endpoints

All three endpoints accept OTLP/HTTP protobuf payloads with Basic auth.

Logs

POST /v1/logs
Content-Type: application/x-protobuf
Authorization: Basic <base64("client_id:api_key")>

OTel log records are mapped to Icelake’s log storage with:

• body → log message
• resource.attributes → labels
• timestamp → nanosecond precision timestamp
• severity → log level label

Metrics

POST /v1/metrics
Content-Type: application/x-protobuf
Authorization: Basic <base64("client_id:api_key")>

OTel metrics are converted to Prometheus-compatible format:

• Gauge → Prometheus gauge
• Sum (monotonic) → Prometheus counter
• Histogram → Prometheus histogram
• Attributes → Prometheus labels

Traces

POST /v1/traces
Content-Type: application/x-protobuf
Authorization: Basic <base64("client_id:api_key")>

Application instrumentation

Direct OTLP export (no Collector)

For simple setups, export directly from your application:

Python Example

import base64
from opentelemetry.exporter.otlp.proto.http.log_exporter import OTLPLogExporter

creds = base64.b64encode(b"your-client-id:ilk_your-api-key-here").decode()
exporter = OTLPLogExporter(
    endpoint="https://api.icelake.eu/v1/logs",
    headers={"Authorization": f"Basic {creds}"},
)

Node.js Example

const { OTLPLogExporter } = require('@opentelemetry/exporter-logs-otlp-http');

const creds = Buffer.from('your-client-id:ilk_your-api-key-here').toString('base64');
const exporter = new OTLPLogExporter({
  url: 'https://api.icelake.eu/v1/logs',
  headers: {
    Authorization: `Basic ${creds}`,
  },
});

Storage layout

OTel data is stored alongside Prometheus and Loki data in Parquet format on S3:

S3 Storage Layout

s3://icelake-data/
├── metrics/    # Prometheus + OTel metrics
├── logs/       # Loki + OTel logs
└── traces/     # OTel traces
    └── tenant=default/
        └── 2026/02/20/
            ├── traces_001.parquet
            └── compacted_001.parquet

API reference

Endpoint	Method	Content-Type	Description
/v1/logs	POST	application/x-protobuf	OTLP log ingestion
/v1/metrics	POST	application/x-protobuf	OTLP metric ingestion
/v1/traces	POST	application/x-protobuf	OTLP trace ingestion

Best practices

Pro tip: use the OpenTelemetry Collector as a gateway to batch, filter, and route data before sending to Icelake.

• Use the batch processor in your OTel Collector for efficient ingestion
• Map resource attributes to meaningful labels for queryability
• Set up separate pipelines for logs, metrics, and traces
• Monitor collector health alongside Icelake ingestion rates

Related docs

• Quickstart — end-to-end Icelake + Grafana setup in 5 minutes
• Prometheus — remote write path for metrics-only pipelines
• Loki & LogQL — log-shipping alternative if you’re not on OTEL