Migrate Traces/APM from LGTM Stack

Overview

This guide walks you through migrating distributed tracing and APM from Tempo (LGTM Stack) to SigNoz. You will:

1. Check your current trace sources in Tempo
2. Redirect your trace data to SigNoz
3. Validate traces are flowing correctly
4. Explore SigNoz's APM capabilities

SigNoz is built natively on OpenTelemetry, making it compatible with most existing instrumentation used with Tempo (OpenTelemetry, Jaeger, Zipkin).

Prerequisites

Before starting, ensure you have:

• A SigNoz account (Cloud) or a running SigNoz instance (Self-Hosted)
• Access to your Grafana/Tempo instance to review current services
• Access to your application source code or collector configuration

Step 1: Assess Your Current Tracing Setup

Before migrating, inventory what you're collecting in Tempo.

List Your Services

In Grafana Explore (select Tempo datasource), you can usually see a list of services in the "Service Name" dropdown of the Search tab. Make a note of:

• Service Names: The applications sending traces.
• Instrumentation: Are they using OpenTelemetry SDKs, Jaeger agents, Zipkin, or Grafana Agent?

Categorize Your Services

Group services by how they're instrumented:

Step 2: Migrate Each Service

Work through each service from your inventory.

From OpenTelemetry SDK

If your applications are already instrumented with OpenTelemetry and sending to Tempo (directly or via a collector), you just need to update the exporter configuration.

For new applications or to switch from other agents, see our Instrumentation Guides.

Before (Tempo/Grafana Agent):

OTEL_EXPORTER_OTLP_ENDPOINT=http://tempo:4317

After (SigNoz Cloud):

OTEL_EXPORTER_OTLP_ENDPOINT=https://ingest.<region>.signoz.cloud:443
OTEL_EXPORTER_OTLP_HEADERS=signoz-ingestion-key=<SIGNOZ_INGESTION_KEY>

Replace:

• <region>: Your SigNoz Cloud region (us, eu, or in)
• <SIGNOZ_INGESTION_KEY>: Your ingestion key from Settings → Ingestion Settings

From Grafana Agent

If you are using Grafana Agent to collect traces, we recommend migrating to the OpenTelemetry Collector. The OTel Collector is the upstream project for Grafana Agent's tracing features and provides the most compatibility.

1. Install OpenTelemetry Collector: Follow the installation guide.
2. Configure Receivers: Copy your receiver configuration (e.g., otlp, jaeger, zipkin) from Grafana Agent to the OTel Collector's config.yaml.
3. Configure Exporter: Set up the OTLP exporter to send to SigNoz.

exporters:
  otlp:
    endpoint: "https://ingest.<region>.signoz.cloud:443"
    headers:
      signoz-ingestion-key: "<SIGNOZ_INGESTION_KEY>"
    tls:
      insecure: false

service:
  pipelines:
    traces:
      receivers: [otlp] # Add jaeger/zipkin if needed
      processors: [batch]
      exporters: [otlp]

From Jaeger Clients

If your applications use legacy Jaeger clients sending to Tempo:

1. Install OpenTelemetry Collector: Installation guide.
2. Enable Jaeger Receiver: Add the Jaeger receiver to your config.yaml.

receivers:
  jaeger:
    protocols:
      grpc:
        endpoint: 0.0.0.0:14250
      thrift_http:
        endpoint: 0.0.0.0:14268

3. Update Pipeline:

service:
  pipelines:
    traces:
      receivers: [jaeger]
      exporters: [otlp]

4. Point Applications: Update your application to send traces to the Collector's Jaeger endpoint (e.g., http://localhost:14268/api/traces).

From Zipkin Clients

If your applications use Zipkin clients:

1. Install OpenTelemetry Collector: Installation guide.
2. Enable Zipkin Receiver: Add the Zipkin receiver to your config.yaml.

receivers:
  zipkin:
    endpoint: 0.0.0.0:9411

3. Update Pipeline:

service:
  pipelines:
    traces:
      receivers: [zipkin]
      exporters: [otlp]

4. Point Applications: Update your application to send traces to the Collector's Zipkin endpoint (e.g., http://localhost:9411/api/v2/spans).

Validate

Verify traces are flowing correctly by comparing against your service inventory.

Check Services Are Reporting

1. In SigNoz, navigate to Services in the left sidebar.
2. Verify each service from your inventory appears in the list.
3. Click a service to see its metrics (latency, error rate, throughput).

Verify Traces

1. Navigate to Traces in the left sidebar.
2. Select a service and time range.
3. Verify spans appear with expected attributes.
4. Check that trace context propagates across service boundaries (parent-child relationships).

Next Steps

Once your traces are flowing to SigNoz:

• Create dashboards with trace-based panels
• Set up trace-based alerts for latency and errors
• Correlate traces with logs for debugging
• Explore the Service Map to understand dependencies
• Migrate your logs next