greptimedb-recipes

CLI tool to install GreptimeDB recipes (flows and dashboards).

Installation

npm install -g greptimedb-recipes

Usage

greptimedb-recipes add <path> [options]

Options

• -e, --env <key=value> - Template variable (can be used multiple times)
• --pg-host <host> - PostgreSQL host (default: "localhost")
• --pg-port <port> - PostgreSQL port (default: "4003")
• --http-host <host> - HTTP API host (default: "localhost")
• --http-port <port> - HTTP API port (default: "4000")
• --dry-run - Show what would be done without executing

Example

# Install a recipe with template variables
greptimedb-recipes add ./opentelemetry-django -e TRACE_TABLE=opentelemetry_traces

# Dry run to see what would be executed
greptimedb-recipes add ./opentelemetry-django --dry-run -e TRACE_TABLE=opentelemetry_traces

# Specify custom GreptimeDB endpoints
greptimedb-recipes add ./opentelemetry-django \
  --pg-host 192.168.1.100 \
  --pg-port 4003 \
  --http-host 192.168.1.100 \
  --http-port 4000

Recipe Structure

A recipe should have the following structure:

my-recipe/
├── dashboard/           # Dashboard JSON files
│   └── my-dashboard.json
├── flow/               # SQL flow files
│   └── my-flow.sql
├── examples/           # Example files (not processed)
│   └── example.py
└── README.md           # Recipe documentation

Flow Directory

SQL files in the flow/ directory are executed against GreptimeDB via the PostgreSQL interface. Template variables can be used with ${VARIABLE_NAME} syntax and replaced using the -e option.

Dashboard Directory

JSON files in the dashboard/ directory are uploaded to GreptimeDB via the HTTP API. The dashboard name is taken from metadata.name in the JSON file.