Pocketdex

Pocketdex is an indexer for the Shannon implementation of the pocket network protocol.

It is built using the SubQuery SDK, which wraps postgraphile to provide a robust GraphQL API to the indexed data.

To learn more about SubQuery, see their docs.

• Usage & Query Docs
  • Explore via postgres
  • Explore via GraphQL
• Getting Started
  • tl;dr local development (if not your first time)
  • 1. Install dependencies
  • 2. Generate types
  • 3. Run
    • Localnet ONLY
    • 3.1 Debugging, errors running & building
    • 3.2 Using a pre-built image
    • 3.3 Available Scripts breakdown
    • 3.4 Using k8s

Usage & Query Docs

See the introduction docs directory for details on how to use indexed data after you're fully set up.

Explore via postgres

Connect to the postgres container, update the schema, and explore!

docker exec -it pocketdex_development-postgres-1 psql -U postgres -d postgres
SET search_path TO app;
\dt

Explore via GraphQL

The poktscan team put together a playground to explore the testnet data.

You can access at shannon-testnet.poktscan.com and use the sample query below to get started.

Click to expand sample query

  query  {
      distinct_poktroll_event_types: events (distinct: TYPE, filter: {type: {includes: "poktroll"}}){
      totalCount
      nodes {
        type
      }
    }

    distinct_poktroll_message_types: messages (distinct: TYPE_URL, filter:{typeUrl: {includes: "poktroll"}}) {
      totalCount
      nodes {
        typeUrl
        # transactionId
      }
    }

    unprocessedEntities (distinct: ERROR) {
      totalCount
      nodes {
        eventId
        messageId
        transactionId
        error
      }
    }

    indexer_metadata: _metadata {
      targetHeight
      lastProcessedHeight
    }


    # tx_events: events(filter: {type: {equalTo: "tx"}}) {
    #   nodes {
    #     attributes {
    #       nodes {
    #         key
    #         value
    #       }
    #     }
    #   }
    # }

    # begin_block_events: events (filter: {attributes: {some: {key: {equalTo: "mode"}, value: {equalTo: "BeginBlock"}}}}) {
    #   nodes {
    #     attributes {
    #       nodes {
    #         key
    #         value
    #       }
    #     }
    #   }
    # }

    # events (distinct: TYPE) {
    #   nodes {
    #     type
    #   }
    # }

    # claims: messages(filter: {typeUrl: {equalTo: "/poktroll.proof.MsgCreateClaim"}}) {
    #   nodes {
    #     typeUrl
    #     json
    #   }
    # }

    # eventAttributes(distinct: KEY) {
    #   nodes {
    #     key
    #   }
    # }
}

Getting Started

tl;dr local development (if not your first time)

Run the following:

yarn install
yarn run codegen
docker context use default # Optional
yarn run docker:build:development
yarn poktroll:proxy:start
yarn run docker:start:development

1. Install dependencies

yarn install

2. Generate types

Types will need to be regenerated any time the graphql.schema is changed.

yarn run codegen

3. Run

Dotenv files will be automatically created after the yarn install thanks to the postinstall script. After that, feel free to modify them as you wish.

You will see three dotenv files, each for the corresponding script and environment:

• .env.production
• .env.development
• .env.test

Alternatively, you can manually create them running:

yarn run env:prepare

For this README we will be running all the commands in development but you can also run them in test or production. Following this structure, you can run every docker command docker:<cmd>:<production|development|test>,

Localnet ONLY

# Run this ONLY IF indexing poktroll localnet.
# This will allows subquery-node to connect with the poktroll validator

# Leave this open in a separated terminal. Interrupting the terminal will stop the container.
yarn poktroll:proxy:start

# To stop and remove the proxy
yarn poktroll:proxy:stop

Build & start:

# Then build docker and start
yarn run docker:build:development
# This will turn on the process under a WATCHER so any change to the project.ts schema.graphql or src will trigger
# the needed builds again.
yarn run docker:start:development

Stop (without deleted data)

yarn run docker:stop:development

Or Stop & clean up (delete postgres data):

yarn run docker:clean:development

3.1 Debugging, errors running & building

If you're hitting errors with the above command, do a nuclear clean of all potential issues:

yarn cache clean
docker builder prune --all
docker context use default

Now pick up from the yarn run docker:build step above.

3.2 Using a pre-built image

If you are unable to build locally, a pre-built image is available on Docker Hub: bryanchriswhite/pocketdex-subquery-node:latest.

To use this image, pull it down:

docker pull bryanchriswhite/pocketdex-subquery-node:latest

Then, re-tag the image to match what docker compose is expecting (assumes the repo root dir is pocketdex):

docker tag bryanchriswhite/pocketdex-subquery-node:latest pocketdex-subquery-node:latest

Alternatively, you may update the docker-compose.<env>.yml file, just remember not to commit this change:

services:
  subquery-node:
    image: bryanchriswhite/pocketdex-subquery-node:latest
    ...

3.3 Available Scripts breakdown

• preinstall - Enforces the use of Yarn as the package manager.
• postinstall - Executes the env:prepare script after the installation process.
• env:prepare - Runs the prepare-dotenv.sh script. It prepares .env files for development, test and production environments using .env.sample.
• codegen - Executes the SubQL Codegen.
• build - Triggers a custom build script written in shell due to its complexity.
• watch:build - Acts similar to the build script, but it also disables the linter for real-time coding with nodemon.
• lint - Executes the source code linter.
• lint:fix - Executes the linter with automatic fixing of solvable issues.
• format - Applies Prettier to the code according to the rules defined in .prettierrc.
• format:ci - Checks the code formatting with Prettier without modifying it.
• vendor:setup - Runs the vendor.sh shell script to prepare, run, and build vendor packages.
• vendor:clean-cache - Removes the .yarn/cache files from vendor packages that can potentially lead to errors.
• vendor:install - Executes the vendor.sh shell script to install dependencies for each vendor package.
• vendor:build - Executes the vendor.sh script to build each vendor package.
• vendor:lint - Executes the vendor.sh script to lint each vendor package.
• vendor:clean - Executes the vendor.sh script to recursively remove node_modules for each vendor package.
• docker:compose - Runs the docker-compose.sh script, which forms the base for all other docker:<action>:<environment> scripts.
• poktroll:proxy:start and poktroll:proxy:stop - Executes proxy-tunnel.sh starting and stopping the proxy tunnel to poktroll localnet validator, respectively.
• docker:check-env:<environment> - Ensures the required .env.<environment> exists by running dotenv-check.sh.
• docker:build:<environment> - Builds docker images for the specified environment.
• docker:build:no-cache:<environment> - Builds docker images for the specified environment without using Docker's cache.
• docker:start:<environment> - Starts all services for the specified environment.
• docker:ps:<environment> - Shows the status of services for the specified environment.
• docker:stop:<environment> - Stops all active services for the specified environment without removing them.
• docker:clean:<environment> - Stops and removes all services, volumes, and networks for the specified environment.

3.4 Using k8s

See the instructions in docs/kubernetes.md for deploying using Kubernetes.