otel-desktop-viewer

otel-desktop-viewer is a CLI tool for receiving OpenTelemetry traces while working on your local machine that helps you visualize and explore your trace data without needing to send it on to a telemetry vendor.

Its goals are to be easy-to-install with minimal dependencies and fast. It is written in Go as a custom exporter on top of the OpenTelemetry Collector. Also, it has a dark mode.

Getting started

Make sure you have go installed.

# install the CLI tool
go install github.com/CtrlSpice/otel-desktop-viewer@latest

# run it!
$(go env GOPATH)/bin/otel-desktop-viewer

# if you have $GOPATH/bin added to your $PATH you can call it directly!
otel-desktop-viewer

# if not you can add it to your $PATH by running this or adding it to
# your startup script (usually ~/.bashrc or ~/.zshrc)
export PATH="$(go env GOPATH)/bin:$PATH"

Running the CLI will open a browser tab to localhost:8000 to load the UI, and spin up a server listening on localhost:4318 for OTLP http payloads and localhost:4317 for OTLP grpc payloads.

Configuring your OpenTelemetry SDK

To send telemetry to otel-desktop-viewer from your application, you need to configure an OTLP exporter to send via grpc to http://localhost:4317 or via http to http://localhost:4318.

If your OpenTelemetry SDK OTLP exporter supports configuration via environment variables then you should be able to send to otel-desktop-viewer with the following environment variables set

# For HTTP:
export OTEL_EXPORTER_OTLP_ENDPOINT="http://localhost:4318"
export OTEL_TRACES_EXPORTER="otlp"
export OTEL_EXPORTER_OTLP_PROTOCOL="http/protobuf"

# For GRPC:
export OTEL_EXPORTER_OTLP_ENDPOINT="http://localhost:4317"
export OTEL_TRACES_EXPORTER="otlp"
export OTEL_EXPORTER_OTLP_PROTOCOL="grpc"

Example with otel-cli

If you have otel-cli installed, you can send some example data with the following script.

# start the desktop viewer (best to do this in a separate terminal)
otel-desktop-viewer

# configure otel-cli to send to our desktop viewer endpoint
export OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4318

# use otel-cli to generate spans!
otel-cli exec --service my-service --name "curl google" curl https://google.com

# a more visually interesting example trace
carrier=$(mktemp)
sockdir=$(mktemp -d)
otel-cli span background \
   --service "otel-cli-example" \
   --name "otel-cli-example background span" \
   --tp-print \
   --tp-carrier $carrier \
   --sockdir $sockdir &
sleep 0.1 # give the background server just a few ms to start up
otel-cli span event --name "cool thing" --attrs "foo=bar" --sockdir $sockdir
otel-cli exec --service "otel-cli-example" --name "curl google" --tp-carrier $carrier curl https://google.com
sleep 0.1
otel-cli exec --service "otel-cli-example" --name "curl google" --tp-carrier $carrier curl https://google.com
sleep 0.1
otel-cli span event --name "another cool thing\!" --attrs "foo=bar" --sockdir $sockdir
otel-cli span end --sockdir $sockdir

Why does this exist?

When we send OpenTelemetry telemetry to a tracing vendor we expect to be able to visualize our data, but when working locally our experience often looks more like this:

{
	"Name": "Poll",
	"SpanContext": {
		"TraceID": "4d7f165e90111f4fc9003d5bbf7aca81",
		"SpanID": "1f7a655a9b7a6f4b",
		"TraceFlags": "01",
		"TraceState": "",
		"Remote": false
	},
	"Parent": {
		"TraceID": "4d7f165e90111f4fc9003d5bbf7aca81",
		"SpanID": "d1b8f7f96ad9d1a0",
		"TraceFlags": "01",
		"TraceState": "",
		"Remote": false
	},
  ...

You can use Jaeger's all-in-one distribution, but this requires quite a bit of additional knowledge for the end-user around docker and navigating a lot of configuration options. Additionally the user experience is not focused around "show me the data I just emitted".

The goals for otel-desktop-viewer are to allow a user to install it with one command, require minimal configuration and additional tooling, and be as approachable as possible for developers at all levels of experience.

Implementation

The CLI is implemented in Go building on top of the OpenTelemetry Collector. A custom desktop exporter is registered that:

• collects trace data in memory
• exposes this trace data via an HTTP API
• serves a static React app that renders the collected traces

All of the static web assets are built into the final binary using the go:embed directive so that the binary is self-contained and relocatable.

What's with the axolotl??

Her name is Lulu Axol'Otel, she is very pink, and I love her.

More seriously, I like to give my side projects an animal theme to add a little aesthetic interest on what otherwise might be fairly plain applications.

License

Apache 2.0, see LICENSE