AWS Athena Client for NodeJS

Installation

Using NPM:

npm install aws-athena-node-client @aws-sdk/client-athena

Using yarn:

yarn add aws-athena-node-client @aws-sdk/client-athena

Using pnpm:

pnpm add aws-athena-node-client @aws-sdk/client-athena

Use

Create client

const athenaNodeClient = new AthenaNodeClient({
  bucketUri: "s3://athena-query-results-eu-west-1/",
  database: "default",
  waitTime: 0.5,
  workGroup: "my-work-group",
});

Run query

const query = `SELECT 1`;

try {
  const results = await athenaNodeClient.executeQuery<T>(query);
  console.log(results);
} catch (error) {
  console.error(error);
}

Cache

If you want to use AWS Athena cache, you can specify that cache in minutes. If you don't provide this query configuration, Athena doesn't use this cache.

const results = await athenaNodeClient.executeQuery<T>(query, {
  cacheInMinutes: 60,
});

Parameters

This project doesn't support native query parameters becase I consider that are very simple and useless, then this project use pg-promise package to format the SQL with parameters to prevent SQL Injection attacks. You can visit https://github.com/vitaly-t/pg-promise?tab=readme-ov-file#named-parameters for more info.

const query =
  "SELECT name, surname, age FROM users WHERE name = $(name) AND surname = $(surname)";

const results = await athenaNodeClient.executeQuery<T>(query, {
  parameters: {
    name: "John",
    surname: "Doe",
  },
});

Run query and get S3 URL with results

You need to install @aws-sdk/client-s3 and @aws-sdk/s3-request-presigner

pnpm add @aws-sdk/client-s3 @aws-sdk/s3-request-presigner

const query = `SELECT 1`;

try {
  const results = await athenaNodeClient.executeQueryAndGetS3Url(query);
  console.log(results); // Print s3://S3_BUCKET_NAME/QUERY_ID.csv
} catch (error) {
  console.error(error);
}

Cancel query

const query = `SELECT 1`;

try {
  const results = await athenaNodeClient.executeQuery<T>(query, {
    parameters,
    id: "hdaiuh33r8uyjdkas",
  });
  console.log(results);
} catch (error) {
  if (!(error instanceof QueryCanceledException)) {
    console.error(error);
  }
}

You must run this code in a distinct thread than the query execution thread.

try {
  await athenaNodeClient.cancelQuery("hdaiuh33r8uyjdkas");
} catch (error) {
  console.error(error);
}

Upgrade from v1 to v2

• Client class renamed to AthenaNodeClient to prevent issues with @aws-sdk/client-athena classes naming.
• Now you must instance AthenaClient from @aws-sdk/client-athena package. This is to optimize support for AWS Lambda functions (@aws-sdk/* packages are included in AWS Lambda runtime), use the most updated version of this package in your project and you can use more custom configurations in the AWS SDK client, like roles. You need to execute:

pnpm add @aws-sdk/client-athena

Before

const athenaClient = new AthenaClient({
  awsConfig: {
    accessKeyId: "DASCDAS82941",
    apiVersion: "2017-05-18",
    region: "eu-west-1",
    secretAccessKey: "CJDADDHDASIUOHADS/3123DASE12",
  },
  bucketUri: "s3://athena-query-results-eu-west-1/",
  database: "default",
  waitTime: 0.5,
  workGroup: "my-work-group",
});

After

import { AthenaClient } from "@aws-sdk/client-athena";

const athena = new AthenaClient({});

const athenaNodeClient = new AthenaNodeClient(athena, {
  bucketUri: "s3://athena-query-results-eu-west-1/",
  database: "default",
  waitTime: 0.5,
  workGroup: "my-work-group",
});

• If you use methods executeQueryAndGetS3Key executeQueryAndGetDownloadSignedUrl now you must create your S3 client before. The reason is the same with the Athena client. You need to execute:

pnpm add @aws-sdk/client-s3 @aws-sdk/s3-request-presigner

Before

const athenaClient = new AthenaClient({
  awsConfig: {
    accessKeyId: "DASCDAS82941",
    apiVersion: "2017-05-18",
    region: "eu-west-1",
    secretAccessKey: "CJDADDHDASIUOHADS/3123DASE12",
  },
  bucketUri: "s3://athena-query-results-eu-west-1/",
  database: "default",
  waitTime: 0.5,
  workGroup: "my-work-group",
});

After

import { AthenaClient } from "@aws-sdk/client-athena";
import { S3 } from "@aws-sdk/client-s3";

const athena = new AthenaClient({});
const s3 = new S3({
  useDualstackEndpoint: true, // recommended to support IPv6
});

const athenaNodeClient = new AthenaNodeClient(athena, {
  bucketUri: "s3://athena-query-results-eu-west-1/",
  database: "default",
  waitTime: 0.5,
  workGroup: "my-work-group",
  s3Client: s3,
});

• Method executeQueryAndGetS3Url is renamed to executeQueryAndGetS3Key and now returns and object with the S3 bucket and key of the object generated by Athena instead of returns an S3 schema url. If you want a direct URL to download the results, you must use executeQueryAndGetDownloadSignedUrl method.

• All query related method, now have 2 parameters, sql and query config. The query parameters and query id, must be inside this second parameter:

Before

const results = await athenaClient.executeQuery<T>(query, parameters, queryId);

After

const results = await athenaNodeClient.executeQuery<T>(query, {
  parameters: {
    name: "John",
    surname: "Doe",
  },
  id: "abcd",
  cacheInMinutes: 60,
});

Upgrade from v2 to v3 and v4

• There is a new statistics parameter in QueryConfig. Now the executeQuery methods will also return a statistics object if the stats parameter is true. By default stats value is false.

athenaNodeClient.executeQuery()

Before

const results = await athenaNodeClient.executeQuery<T>(query);

• V2 Response example:

[
  {
    "result": "1"
  },
  {
    "result": "2"
  }
]

After

• If the statistics parameter is true, the method returns an object with results and statistics, otherwise it will return an array of objects.

const result = await athenaNodeClient.executeQuery<T>(query, {
  stats: false,
});
// OR
const { results, statistics } = await athenaNodeClient.executeQuery<T>(query, {
  stats: true,
});

• V3 response example with stats false or without stats param:

{
  "id": "id",
  "results": [
    {
      "result": "1"
    },
    {
      "result": "2"
    }
  ]
}

• V3 response example with stats true:

{
  "id": "id",
  "results": [
    {
      "result": "1"
    },
    {
      "result": "2"
    }
  ],
  "statistics": {
    "dataScannedInBytes": 26376236,
    "executionTimeInSeconds": 2.703
  }
}

athenaNodeClient.executeQueryAndGetS3Key()

Before

const { bucket, key } = await athenaNodeClient.executeQueryAndGetS3Key<T>(
  query
);

• V2 Response example:

{
  "bucket": "bucket-test",
  "key": "123-key.csv"
}

After

If the statistics parameter is true, the method returns an object with bucket, key and statistics, otherwise it will return an object with bucket and key.

const { bucket, key, statistics } =
  await athenaNodeClient.executeQueryAndGetS3Key<T>(query, {
    stats: true | false,
  });

• V3 response example with stats false or without stats param:

{
  "id": "id",
  "bucket": "bucket-test",
  "key": "123-key.csv"
}

• V3 response example with stats true:

{
  "id": "id",
  "bucket": "bucket-test",
  "key": "123-key.csv",
  "statistics": {
    "dataScannedInBytes": 26376236,
    "executionTimeInSeconds": 2.703
  }
}

athenaNodeClient.executeQueryAndGetDownloadSignedUrl()

Before

const url = await athenaNodeClient.executeQueryAndGetDownloadSignedUrl<T>(
  query
);

• V2 response example:

"https://bucket-test..."

After

If the statistics parameter is true, the method returns an object with url and statistics, otherwise it will return an object with url.

const { url, statistics } =
  await athenaNodeClient.executeQueryAndGetDownloadSignedUrl<T>(query, {
    stats: true | false,
  });

• V3 response example with stats false or without stats param:

{
  "id": "id",
  "url": "https://bucket-test..."
}

• V3 response example with stats true:

{
  "id": "id",
  "url": "https://bucket-test...",
  "statistics": {
    "dataScannedInBytes": 26376236,
    "executionTimeInSeconds": 2.362
  }
}