Akash Quick Start

SubQuery TeamAbout 2 min

Akash Quick Start

The goal of this quick start guide is to index all reward transactionsopen in new window for delegators in the Akash networkopen in new window.

Info

This network is based on the Cosmos SDK, which means you can index chain data via the standard Cosmos RPC interface.

Before we begin, make sure that you have initialised your project using the provided steps in the Start Here section. You must complete the suggested 4 stepsopen in new window for Cosmos users.

Tips

The final code of this project can be found hereopen in new window.

Your Project Manifest File

The Project Manifest file is an entry point to your project. It defines most of the details on how SubQuery will index and transform the chain data.

For Cosmos chains, there are four types of mapping handlers (and you can have more than one in each project):

BlockHanders: On each and every block, run a mapping function
TransactionHandlers: On each and every transaction, run a mapping function
MessageHandlers: On each and every message that matches optional filter criteria, run a mapping function
EventHanders: On each and every event that matches optional filter criteria, run a mapping function

Note that the manifest file has already been set up correctly and doesn’t require significant changes, but you need to change the datasource handlers. This section lists the triggers that the manifest file looks for on the blockchain to start indexing.

{
  dataSources: [
    {
      kind: CosmosDatasourceKind.Runtime,
      startBlock: 11364001,
      mapping: {
        file: "./dist/index.js",
        handlers: [
          {
            handler: "handleReward",
            kind: CosmosHandlerKind.Event,
            filter: {
              type: "withdraw_rewards",
              messageFilter: {
                type: "/cosmos.distribution.v1beta1.MsgWithdrawDelegatorReward",
              },
              /*
                contractCall field can be specified here too
                values: # A set of key/value pairs that are present in the message data
                contract: "juno1v99ehkuetkpf0yxdry8ce92yeqaeaa7lyxr2aagkesrw67wcsn8qxpxay0"
              */
            },
          },
        ],
      },
    },
  ],
}

In the code above, we have defined a single handler, handleReward, that will be executed whenever a withdraw_rewards type is detected within a MsgWithdrawDelegatorReward type message. This handler is used to track the rewards transactions of delegators in the Akash network.

Check out our Manifest File documentation to get more information about the Project Manifest (project.ts) file.

Update Your GraphQL Schema File

The schema.graphql file determines the shape of your data from SubQuery due to the mechanism of the GraphQL query language. Hence, updating the GraphQL Schema file is the perfect place to start. It allows you to define your end goal right at the start.

For this project, you'll need to modify your schema.graphql file as follows. Since we're indexing all reward transactionsopen in new window for delegators in the Akash network, we have a DelegatorReward entity that comprises a number of properties, including reward amount, delegator information, validator's address, and so forth. We also have a Delegator entity, which keeps track of the total rewards of each delegator.

type DelegatorReward @entity {
  id: ID!
  blockHeight: BigInt
  txHash: String
  feeDenomination: String
  feeAmount: BigInt
  rewardAmount: BigInt!
  delegator: Delegator!
  validatorAddress: String
}

type Delegator @entity {
  id: ID!
  totalRewards: BigInt!
}

Note

Importantly, these relationships can not only establish one-to-many connections but also extend to include many-to-many associations. To delve deeper into entity relationships, you can refer to this section. If you prefer a more example-based approach, our dedicated Hero Course Module can provide further insights.

SubQuery simplifies and ensures type-safety when working with GraphQL entities, actions, and transactions.

yarn

yarn codegen

npm

npm run-script codegen

This action will generate a new directory (or update the existing one) named src/types. Inside this directory, you will find automatically generated entity classes corresponding to each type defined in your schema.graphql. These classes facilitate type-safe operations for loading, reading, and writing entity fields. You can learn more about this process in the GraphQL Schema section.

If you've expressed a preference to employ the Cosmos message based on the provided proto files, this command will also generate types for your listed protobufs and save them into src/types directory, providing you with more typesafety. For example, you can find Osmosis' protobuf definitions in the official documentationopen in new window. Read about how this is done in Cosmos Codegen from CosmWasm Protobufs and Cosmos Manifest File Configuration.

Now that you have made essential changes to the GraphQL Schema file, let’s go ahead with the next configuration.

Check out the GraphQL Schema documentation to get in-depth information on schema.graphql file.

Now that you have made essential changes to the GraphQL Schema file, let’s proceed ahead with the Mapping Function’s configuration.

Add a Mapping Function

Mapping functions define how blockchain data is transformed into the optimised GraphQL entities that we previously defined in the schema.graphql file.

Navigate to the default mapping function in the src/mappings directory and update your mapping files to match the following (note the additional imports):

import assert from "assert";
import { DelegatorReward, Delegator } from "../types";
import {
  CosmosEvent,
  CosmosBlock,
  CosmosMessage,
  CosmosTransaction,
} from "@subql/types-cosmos";

export async function handleReward(event: CosmosEvent): Promise<void> {
  logger.info(
    `New Reward Withdraw event at block ${event.block.header.height.toString()}`,
  );

  const recordAmountString = event.event.attributes.find(
    (a) => a.key === "amount",
  )?.value;
  const validatorAddress = event.event.attributes.find(
    (a) => a.key === "validator",
  )?.value;
  const delegatorAddress = event.tx.tx.events
    .find((e) => e.type === "coins_received")
    ?.attributes.find((a) => a.key === "receiver")?.value;

  var rewardBigInt = BigInt(0);
  if (recordAmountString && Array.from(recordAmountString)[0] != "0") {
    rewardBigInt = BigInt(recordAmountString.split("u")[0]);
  }

  // Confirm we have all required values
  assert(
    rewardBigInt && validatorAddress && delegatorAddress,
    "No reward or no msg found",
  );

  const rewardRecord = DelegatorReward.create({
    id: event.tx.hash,
    blockHeight: BigInt(event.block.block.header.height),
    txHash: event.tx.hash,
    feeDenomination: event.tx.decodedTx.authInfo.fee?.amount[0].denom,
    feeAmount: event.tx.decodedTx.authInfo.fee?.amount[0].amount
      ? BigInt(event.tx.decodedTx.authInfo.fee?.amount[0].amount)
      : undefined,
    rewardAmount: rewardBigInt,
    delegatorId: delegatorAddress,
    validatorAddress: validatorAddress,
  });

  await handleDelegator(rewardRecord.rewardAmount, rewardRecord.delegatorId);
  await rewardRecord.save();
}

async function handleDelegator(reward: bigint, delegatorAddress: string) {
  let delegatorRecord = await Delegator.get(delegatorAddress);

  if (!delegatorRecord) {
    // Create new Delegator record
    delegatorRecord = Delegator.create({
      id: delegatorAddress,
      totalRewards: reward,
    });
    logger.info(
      "New delegator" + delegatorAddress + "with reward" + reward.toString(),
    );
  } else {
    // Update delegators total
    delegatorRecord.totalRewards = delegatorRecord.totalRewards + reward;
  }
  await delegatorRecord.save();
}

In the Akash SubQuery project, we have two main functions, namely handleReward and handleDelegator, that were defined in the src/mappings/mappingHandlers.ts file.

The handleReward function is triggered when a withdraw_rewards type event is detected. It receives an event of type CosmosEvent, logs a message to the console for debugging purposes, and then tries to extract key data points such as the reward amount, validator address, and delegator address from the event.event.attributes (Cosmos events code attributes as an array of key-value pairs).

It then uses these attributes to create a DelegatorReward object, recording the reward amount and validator address associated with each reward withdrawal event. This object is subsequently saved in the GraphQL entities that we previously defined in the schema.graphql file.

The handleDelegator function is invoked within the handleReward function. This function is responsible for managing the Delegator entity. If a delegator does not already exist in the GraphQL entities, a new one is created with the reward from the current transaction. If the delegator already exists, the total rewards for the delegator are updated by adding the current reward.

This way, we're able to track all delegator rewards on the Akash network, along with the validator from whom the reward came. It's crucial to note that the handleDelegator function handles the Delegator entity creation and updates, whilst the handleReward function creates the DelegatorReward entity.

Check out our Mappings documentation and get information on the mapping functions in detail.

Build Your Project

Next, build your work to run your new SubQuery project. Run the build command from the project's root directory as given here:

yarn

yarn build

npm

npm run-script build

Important

Whenever you make changes to your mapping functions, you must rebuild your project.

Now, you are ready to run your first SubQuery project. Let’s check out the process of running your project in detail.

Whenever you create a new SubQuery Project, first, you must run it locally on your computer and test it and using Docker is the easiest and quickiest way to do this.

Run Your Project Locally with Docker

The docker-compose.yml file defines all the configurations that control how a SubQuery node runs. For a new project, which you have just initialised, you won't need to change anything.

However, visit the Running SubQuery Locally to get more information on the file and the settings.

Run the following command under the project directory:

yarn

yarn start:docker

npm

npm run-script start:docker

Note

It may take a few minutes to download the required images and start the various nodes and Postgres databases.

Query your Project

Next, let's query our project. Follow these three simple steps to query your SubQuery project:

Open your browser and head to http://localhost:3000.
You will see a GraphQL playground in the browser and the schemas which are ready to query.
Find the Docs tab on the right side of the playground which should open a documentation drawer. This documentation is automatically generated and it helps you find what entities and methods you can query.

Try the following queries to understand how it works for your new SubQuery starter project. Don’t forget to learn more about the GraphQL Query language.

{
  query {
    delegatorReward(first: 1) {
      nodes {
        id
        delegatorAddress
        validatorAddress
        rewardAmount
      }
    }
  }
}

You will see the result similar to below:

{
  "data": {
    "query": {
      "delegatorRewards": {
        "totalCount": 1,
        "nodes": [
          {
            "id": "808FED7F3FE680EEF8E005EC1927C0CF00D2975E4B26CEE7A098D5DA7DEA8217",
            "delegatorAddress": "akash1ckc4z334rfpdcqdz5a85mdtaktt2dtkm58783c5",
            "validatorAddress": "akashvaloper14kn0kk33szpwus9nh8n87fjel8djx0y0uzn073",
            "rewardAmount": "1440293uakt"
          }
        ]
      }
    }
  }
}

What's next?

Congratulations! You have now a locally running SubQuery project that accepts GraphQL API requests for transferring data.

Tip

Find out how to build a performant SubQuery project and avoid common mistakes in Project Optimisation.

Click here to learn what should be your next step in your SubQuery journey.