我的CPU在Google Cloud上的使用率已达到160％，我似乎不知道为什么吗？

Question

So I was working on this project and I've deployed it to the virtual machine on google cloud with a machine type of g1-small (v1 CPU, 1.7 GB). In the monitoring section, the CPU reaches upwards of 175% with 160% being the average. Google recommends me to increase my machine type to a custom (v1 CPU, 1.75 GB), but I'm not quite sure what's eating up the CPU.

I have read into blocking and non-blocking in node.js and I know that my code is currently non-blocking so that may be an issue. But if I change it to blocking, the code will run sequentially and it will run slower.

Thank you in advance.

Here's my code: https://github.com/QuaziH/CUNYfirst-API/blob/master/looper.js

const db = require('./db/index');
const text = require('./text');
const api = require('./cf-api');

let looper = () => {
    db.query(`SELECT DISTINCT institution, term, subject FROM classes`, async (error, response) => {
        if (error) {
            return console.error('Error fetching client', error);
        }

        if(response.rowCount === 0){
            console.log(`Table is empty`);
        } else {
            for (let i = 0; i < response.rows.length; i++) {
                let allClasses = await api.getAllSections(`${response.rows[i].institution}`, `${response.rows[i].term}`, `${response.rows[i].subject}`);
                db.query(`SELECT phone, carrier, topic, class_num FROM classes WHERE institution='${response.rows[i].institution}' AND term='${response.rows[i].term}' AND subject='${response.rows[i].subject}'`, (error, response) => {
                    if (error) {
                        return console.error('Error fetching client', error);
                    }

                    for (let j = 0; j < response.rows.length; j++) {
                        let status;
                        try {
                            status = allClasses[response.rows[j].class_num].Status;
                        } catch (e) {
                            if (response.rows[j].carrier === '@tmomail.net' || response.rows[j].carrier === '@mymetropcs.com') {
                                text.emailError(`${response.rows[j].phone}${response.rows[j].carrier}`, `Your class ${response.rows[j].topic}- ${response.rows[j].class_num} does not exist in CUNYfirst anymore`);
                            } else {
                                text.twilio(`${response.rows[j].phone}`, `Uh oh! \n Your class ${response.rows[j].topic}- ${response.rows[j].class_num} does not exist in CUNYfirst anymore`);
                            }
                            db.query(`DELETE FROM classes WHERE phone='${response.rows[j].phone}' AND class_num=${response.rows[j].class_num}`, (err, res) => {
                                if (err) {
                                    console.error(`Error deleting from database: ${err}`);
                                }
                            });
                            continue;
                        }

                        if (status === 'Closed') {
                            console.log(`Class is still closed ${response.rows[j].topic}- ${response.rows[j].class_num}`);
                        } else if (status === 'Open') {
                            if (response.rows[j].carrier === '@tmomail.net' || response.rows[j].carrier === '@mymetropcs.com') {
                                text.emailOpen(`${response.rows[j].phone}${response.rows[j].carrier}`, `Your class ${response.rows[j].topic}- ${response.rows[j].class_num} is now open!`);
                            } else {
                                text.twilio(`${response.rows[j].phone}`, `Class Opened! \n Your class ${response.rows[j].topic}- ${response.rows[j].class_num} is now open!`);
                            }
                            db.query(`DELETE FROM classes WHERE phone='${response.rows[j].phone}' AND class_num=${response.rows[j].class_num}`, (err, res) => {
                                if (err) {
                                    console.error(`Error deleting from database: ${err}`);
                                }
                            });
                        }
                    }
                });
            }
        }
    });
};

setInterval(looper, 300000);

Answer 1

Well that explains because Google Cloud Platform uses virtual machines. Virtual machines want to perform like a bare metal machine. My next step would be to analyze your Cpu performance. You can do it from opening two SSH windows (one launching the code, other monitoring) or by connecting from your remote. From the Google Cloud you have tools like the Stackdriver monitoring (which would be more than suitable for this testing) and the Stackdriver login . Also you have the serial port output from your instance. With this information you can make the changes to adapt this instance to yoru code.