Executing a Query (Async)

The Snowflake SDK supports asynchronous query execution. This comparison shows how much simpler this becomes with Promise-based code.

Two examples demonstrate different async patterns, both based on Snowflake SDK documentation.

First Example: Basic Async Execution

Without this library: Complex manual Promise wrapping, callback coordination, query ID management, and manual result streaming.

With this library: Execute the query and await the results.

Second Example: Async with Status Polling

Without this library: Manual Promise wrapping, callback coordination, and complex polling setup.

With this library: Execute with async option, await statement readiness, then implement clean polling.

Examples

Example 1

tip

• Click "WITHOUT Helper Library" to see the original code
• Click "WITH Helper Library" to see the improved code using this library

WITHOUT Helper Library

Based on an example from the Snowflake documentation.

// Not using the Promise Helper Library

let queryId;

// 1. Manually wrap execution to capture query ID
await new Promise((resolve) => {
  connection.execute({
    sqlText: "CALL SYSTEM$WAIT(3, 'SECONDS')",
    asyncExec: true,
    complete: function (err, stmt, rows) {
      if (err) {
        console.error("Query failed:", err);
        return;
      }
      queryId = stmt.getQueryId();
      resolve();
    },
  });
});

// 2. Manually get results and set up streaming
const statement = await connection.getResultsFromQueryId({ queryId: queryId });
await new Promise((resolve, reject) => {
  const stream = statement.streamRows();
  stream.on("error", function (err) {
    reject(err);
  });
  stream.on("data", function (row) {
    console.log(row);
  });
  stream.on("end", function () {
    resolve();
  });
});

WITH Helper Library

// Using the Promise Helper Library

// With this library, you get async behavior by default
// No need for asyncExec: true unless you need immediate query ID access
const { resultsPromise } = connection.execute({
  sqlText: "CALL SYSTEM$WAIT(3, 'SECONDS')",
});

const results = await resultsPromise;
console.log(results);

Example 2

tip

• Click "WITHOUT Helper Library" to see the original code
• Click "WITH Helper Library" to see the improved code using this library

WITHOUT Helper Library

Based on an example from the Snowflake documentation.

// Not using the Promise Helper Library

let queryId;

// 1. Manually wrap execution to capture query ID
await new Promise((resolve, reject) => {
  const statement = connection.execute({
    sqlText: "CALL SYSTEM$WAIT(3, 'SECONDS')",
    asyncExec: true,
    complete: function (err, stmt, rows) {
      if (err) {
        reject(err);
        return;
      }
      queryId = stmt.getQueryId();
      resolve();
    },
  });
});

// 2. Manual polling with setTimeout Promise wrapper
const seconds = 2;
let status = await connection.getQueryStatus(queryId);
while (connection.isStillRunning(status)) {
  console.log(`Query status is ${status}, waiting ${seconds} seconds`);

  await new Promise((resolve) => {
    setTimeout(() => resolve(), 1000 * seconds);
  });

  status = await connection.getQueryStatus(queryId);
}

console.log(`Query has finished executing, status is ${status}`);

WITH Helper Library

// Using the Promise Helper Library

import { scheduler } from "node:timers/promises";

const { statement, resultsPromise } = connection.execute({
  sqlText: "CALL SYSTEM$WAIT(3, 'SECONDS')",
  asyncExec: true,
});

// With asyncExec, await to ensure the statement is ready
await resultsPromise;
const queryId = statement.getQueryId();

// Clean polling with modern timer API
let status = await connection.getQueryStatus(queryId);
while (connection.isStillRunning(status)) {
  console.log(`Query status is ${status}, waiting 2 seconds`);
  await scheduler.wait(2000);
  status = await connection.getQueryStatus(queryId);
}

console.log(`Query has finished executing, status is ${status}`);