callbackWaitsForEmptyEventLoop and cached connections/pools?

Serverless Architectures
1

I read a few articles on calling rds from lambdas and how to use ‘context.callbackWaitsForEmptyEventLoop = false’ in order to try to hand on to cached connections in warm lambdas.

I have read some that advise to use a connection pool of size 1, but is there even any point in using a conenciton pool. Wouldn’t a simple connection like below serve the exact same purpose ?

import * as dynamoDbLib from "../../libs/dynamodb-lib";
import { success, failure } from "../../libs/response-lib";

var mysql = require('mysql2/promise');

let cachedConnection = null

export const main = async (event, context, callback) => {

    console.log('event', JSON.stringify(event))
    try {

        context.callbackWaitsForEmptyEventLoop = false;

        if (cachedConnection === null) {
            console.log('creating connection')
            cachedConnection = await mysql.createConnection({
                host: "dev.xxxx.rds.amazonaws.com",
                user: "xxxxxxxx",
                password: "xxxxxxxxxxx",
                database: "orders-service",
            });
            console.log('connection created')
        } else {
            console.log('using cached connection')
        }

        const [rows, fields] = await cachedConnection.query('insert into test set ?', { orderId: event.id});

        callback(null, success(rows));

    } catch (error) {
        console.log(error)
        callback(null, failure({ status: false, error: error }));
    }
};
2

I was wondering the same. Any further insight on this?