Task Executor

GitHub npm node-current npm type definitions

Task Executor

What's TaskExecutor?

TaskExecutor facilitates building of applications that utilize the computational power of the Golem Network in a transparent and efficient manner. It is a @golem-sdk/golem-js based library allowing running computation tasks, designed for batch map-reduce like scenarios.

With TaskExecutor, developers can focus on implementing their computational tasks without delving into the details of communicating with the Golem Network or managing modules such as payments or market.

System requirements

To use task-executor, it is necessary to have yagna installed, with a recommended minimum version of v0.15.0. Yagna is a service that communicates and performs operations on the Golem Network, upon your requests via the SDK. You can follow these instructions to set it up.

Simplified installation steps

In order to get started and on Golem Network and obtain test GLM tokens (tGLM) that will allow you to build on the test network, follow these steps:

Join the network as a requestor and obtain test tokens

# Join the network as a requestor
curl -sSf https://join.golem.network/as-requestor | bash -

# Start the golem node on your machine,
# you can use `daemonize` to run this in background
yagna service run

# IN SEPARATE TERMINAL (if not daemonized)
# Initialize your requestor
yagna payment init --sender --network holesky

# Request funds on the test network
yagna payment fund --network holesky

# Check the status of the funds
yagna payment status --network holesky

Obtain your `app-key` to use with TaskExecutor

If you don't have any app-keys available from yagna app-key list, go ahead and create one with the command below. You will need this key in order to communicate with yagna from your application via golem-js.You can set it as YAGNA_APPKEY environment variable.

yagna app-key create my-golem-app

Installation

@golem-sdk/task-executor is available as a NPM package.

npm install @golem-sdk/task-executor

Building

To build a library available to the Node.js environment:

npm run build

This will generate production code in the dist/ directory ready to be used in your Node.js or browser applications.

Usage

Hello World example

import { TaskExecutor } from "@golem-sdk/task-executor";
import { pinoPrettyLogger } from "@golem-sdk/pino-logger";

(async function main() {
  const executor = await TaskExecutor.create({
    logger: pinoPrettyLogger({ level: "info" }),
    demand: {
      workload: {
        imageTag: "golem/alpine:latest",
      },
    },
    market: {
      rentHours: 0.5,
      pricing: {
        model: "linear",
        maxStartPrice: 0.5,
        maxCpuPerHourPrice: 1.0,
        maxEnvPerHourPrice: 0.5,
      },
    },
  });
  try {
    await executor.run(async (exe) => console.log((await exe.run("echo 'Hello World'")).stdout));
  } catch (error) {
    console.error("Computation failed:", error);
  } finally {
    await executor.shutdown();
  }
})();

More examples

The examples directory in the repository contains various usage patterns for the TaskExecutor. You can browse through them and learn about the recommended practices. All examples are automatically tested during our release process.

In case you find an issue with the examples, feel free to submit an issue report to the repository.

You can find even more examples and tutorials in the JavaScript API section of the Golem Network Docs.

Supported environments

The library is designed to work with LTS versions of Node (starting from 18) and with browsers.

Documentation

Learn about the Task Executor basic building block - the task function in the Task Model documentation
Learn the Golem Market Basics to optimize your interactions with the Providers

Debugging

The library uses the debug package to provide debug logs. To enable them, set the DEBUG environment variable to task-executor:* to see the related log lines. For more information, please refer to the debug package documentation.

Testing

Read the dedicated testing documentation to learn more about how to run tests of the library.

Contributing

It is recommended to run unit tests and static code analysis before committing changes.

npm run lint
# and
npm run format