kaspa-graph-inspector

KGI is comprised of four components:

• A postgres database
• A processing kaspa node (this is simply a kaspad wrapped in some extra logic)
• An api REST server
• A web server

How the components interact:

• The processing node connects to the Kaspa network the same way a regular kaspad node does and starts syncing just as a kaspad node would
• While it's syncing, it writes metadata about every block to the postgres database
• From the other end, the web server listens to http requests on some port
• When a user navigates their browser to that port, the web server serves the KGI clientside logic, which includes the UI
• The clientside logic calls the api REST server every so often
• The api REST server queries the postgres database and returns it to the clientside
• The clientside uses the response it received from the api REST server to update the UI

Development

For development, it's recommended to run KGI from within Docker

1. Make sure you have docker installed by running docker --version
2. Make sure you have docker-compose installed by running docker-compose --version
3. Define the following environment variables:
   1. POSTGRES_USER=username
   2. POSTGRES_PASSWORD=password
   3. POSTGRES_DB=database-name
   4. API_ADDRESS=localhost
   5. KATNIP_ADDRESS=localhost
   6. API_PORT=4575
   7. WEB_PORT=8080
   8. KASPAD_VERSION=4a560f25a60e876b58d2643ca6eb7e07525e76cc (this can be either a specific kaspda commit hash or a kaspad tag)
   9. KASPA_LIVE_ADDRESS=localhost
4. Run: ./docker-run.sh

Deployment

1. Deploy a postgres database instance in any way you desire. Note the address, port, username, password, and database name, since these will be required later
2. Build processing
   1. Make sure the go build environment is set up by running go version
   2. Within the processing directory, edit go.mod:
      1. Delete the line that starts with replace github.com/kaspanet/kaspad
      2. Set your desired kaspad version in the line under require that starts with github.com/kaspanet/kaspad
   3. Within the processing directory, run go build -o kgi-processing .. This will produce an executable file named kgi-processing
   4. Copy kgi-processing and database directory (also within the processing directory) to wherever you wish to run the node from
3. Build api
   1. Make sure the nodejs build environment is set up by running npm version
   2. Within the api directory, run: npm install
   3. Copy the entire api directory to wherever you wish to run the api server from
4. Build web
   1. Make sure the nodejs build environment is set up by running npm version
   2. Within the web directory, run: npm install
   3. Set the following environment variables:
      1. REACT_APP_API_ADDRESS=example.com:1234 (this is the public address of where your api server will be)
      2. REACT_APP_EXPLORER_ADDRESS=explorer.kaspa.org
      3. REACT_APP_KASPA_LIVE_ADDRESS=kaspa.live
   4. Within the web directory, run: npm run build
   5. Copy the entire web directory to wherever you wish to run the web server from
5. Run processing
   1. Navigate to wherever you copied kgi-processing and database to
   2. Set the following environment variables:
      1. POSTGRES_USER=username
      2. POSTGRES_PASSWORD=password
      3. POSTGRES_DB=database-name
      4. POSTGRES_HOST=database.example.com
      5. POSTGRES_PORT=5432
   3. Run: kgi-processing --connection-string=postgres://${POSTGRES_USER}:${POSTGRES_PASSWORD}@${POSTGRES_HOST}:${POSTGRES_PORT}/${POSTGRES_DB}?sslmode=disable
6. Run api
   1. Navigate to wherever you copied api to
   2. Run: npm run start
7. Run web
   1. Navigate to wherever you copied web to
   2. Run: npm install -g serve
   3. Set the WEB_PORT environment variable to the port you wish to serve the KGI UI from
   4. Run: serve --listen=${WEB_PORT}