Telemetry generator

Prerequisites & setup

Non-development Workflow: Run a published image

If you just want to run (vs build or develop), you can run the most recent image published to this repository's container registry.

For defaults, see Dockerfile.

$ export LS_ACCESS_TOKEN=your token
# can override to any other OTLP endpoint
$ export OTEL_EXPORTER_OTLP_TRACES_ENDPOINT=ingest.lightstep.com:443
$ docker run -e LS_ACCESS_TOKEN --rm ghcr.io/lightstep/telemetry-generator:latest

OpenTelemetry collector builder

Install the OpenTelemetry Collector Builder:

1. $ go install go.opentelemetry.io/collector/cmd/[email protected]

Get the code

1. Clone the telemetry generator repo to a directory of your choosing:
   1. $ cd ~/Code (or wherever)
   2. $ git clone https://github.com/lightstep/telemetry-generator
   3. $ cd telemetry-generator
2. Copy hipster_shop.yaml to dev.yaml for local development. Not strictly necessary (you can point the TOPO_FILE environment variable to any config file) but will potentially save heartache and hassle 😅 This file is in .gitignore, so it won't be included in your commits. If you want to share config changes, add them to a new example config file. $ cp examples/hipster_shop.yaml examples/dev.yaml

Environment variables

• LS_ACCESS_TOKEN = Access token used for sending DEMO telemetry
• LS_ACCESS_TOKEN_INTERNAL = Access token used for sending META (self monitoring) telemetry
• OTEL_EXPORTER_OTLP_TRACES_ENDPOINT = Endpoint for ingesting DEMO telemetry
• OTEL_EXPORTER_OTLP_TRACES_ENDPOINT_INTERNAL = Endpoint for ingesting META telemetry.

Access token

To send demo telemetry data to Lightstep, you'll need an access token associated with the lightstep project you want to use. Go to ⚙ -> Access Tokens to copy an existing one or create a new one. Then:

$ export LS_ACCESS_TOKEN="<your token>"
$ export LS_ACCESS_TOKEN_INTERNAL="<your token>"

Collector endpoint

The env var OTEL_EXPORTER_OTLP_TRACES_ENDPOINT determines the endpoint for demo traces and metrics. To send data to Lightstep, use:

$ export OTEL_EXPORTER_OTLP_TRACES_ENDPOINT=ingest.lightstep.com:443
$ export OTEL_EXPORTER_OTLP_TRACES_ENDPOINT_INTERNAL=ingest.lightstep.com:443

Topo file (generatorreceiver config)

The env var TOPO_FILE determines which config file the generatorreceiver uses.

If you're using the builder you'll want to point to examples/<filename.yaml>:

$ export TOPO_FILE=examples/dev.yaml

For Docker builds, these files are copied to /otel/examples, so set TOPO_FILE like this:

$ export TOPO_FILE=/otel/examples/dev.yaml

Development Workflows

These steps build the collector from the source in this repo.

Build and run the collector

There are two options here, but if possible we recommend using the OpenTelemetry Collector Builder, which is much faster and lets you test config changes without rebuilding. With the Docker build method, you need to rebuild the image for all changes, code or config, and the build process takes much longer.

1. Build and run with the OpenTelemetry Collector Builder (recommended)

(You must first install the builder; see Prerequisites above.)

# Running this local will output the collector Go files into dist/
$ make build
$ make run-local

When using the builder, you only need to re-run the first command for code changes; for config changes just re-run the second command. To run with a different topo file, change the TOPO_FILE environment variable.

If you run into errors while building, please open an issue.

2. Build and run with Docker (alternative)

$ docker build -t lightstep/local-telemetry-generator:latest .
$ export LS_ACCESS_TOKEN=<access-token-for-demo-telemetry>
$ export LS_ACCESS_TOKEN_INTERNAL=<access-token-for-meta-telemetry>
$ export OTEL_EXPORTER_OTLP_TRACES_ENDPOINT=<ingest-endpoint-for-demo-telemetry>
$ export OTEL_EXPORTER_OTLP_TRACES_ENDPOINT_INTERNAL=<ingest-endpoint-for-meta-telemetry>
$ docker run --rm -e LS_ACCESS_TOKEN -e LS_ACCESS_TOKEN_INTERNAL -e OTEL_EXPORTER_OTLP_TRACES_ENDPOINT -e OTEL_EXPORTER_TRACES_ENDPOINT_INTERNAL --env TOPO_FILE=/etc/otel/hipster_shop.yaml lightstep/local-telemetry-generator:latest

When building with Docker, you need to re-run both steps for any code or config changes. If you run into errors while building, please open an issue.

Publishing a Release

These steps enable a new Docker image to be available with docker pull ghcr.io/lightstep/telemetry-generator:<tag>

1. Make your code changes and add to a new PR, ensure to include an:
   • Update to VERSION in the file VERSION
   • Update to CHANGELOG.md
   • Update to Compatibility Matrix below.
2. Create PR, get approvals, merge changes
3. Run make add-tag
   • (This will run git tag under the hood using the version number in VERSION)
4. Run make push-tag
   • (This will push the tags to Github. THIS is the operation that will kick off the GHA workflow, build and push a new image out to GHCR.io)

Compatibility Matrix

Telemetry generator should be built with the compatible open-telemetry Collector builder binary, with collector and collector-contrib components of the same version. Below is a matrix showing the correct collector version for the 10 most recent telemetry-generator versions.

Telemetry Generator	OpenTelemetry Collector
v0.15.0	v0.88.0
v0.14.2	v0.83.0
v0.14.1	v0.83.0
v0.14.0	v0.83.0
v0.13.0	v0.83.0
v0.12.0	v0.78.1
v0.11.13	v0.69.1
v0.11.12	v0.69.1
v0.11.11	v0.69.1
v0.11.10	v0.69.1
v0.11.9	v0.68.0
v0.11.8	v0.67.0
v0.11.7	v0.60.0
v0.11.6	v0.60.0
v0.11.5	v0.60.0
v0.11.4	v0.60.0
v0.11.3	v0.60.0
v0.11.2	v0.60.0
v0.11.1	v0.60.0
v0.11.0	v0.60.0