How to Build a Visual MongoDB Slow Query Dashboard with PHP

Overview

MongoDB stores profiling data for slow operations in the system.profile collection. This tool is a PHP web application that periodically extracts slow queries from MongoDB, stores them in a MySQL database, and presents them via a browser UI.

Prerequisites

PHP with php-pear, php-devel, php-mysql and the legacy mongo extension (or compatible driver).

MySQL server for result storage.

MongoDB version 3.2 or newer (2.x not tested).

Environment Setup

1. Install PHP‑MySQL driver

yum install -y php-pear php-devel php gcc openssl openssl-devel cyrus-sasl cyrus-sasl-devel httpd mysql php-mysql

2. Install PHP‑Mongo driver

pecl install mongo

Add extension=mongo.so to /etc/php.ini and restart Apache:

service httpd restart

3. Create a MongoDB monitoring user

The collector must be able to run db.runCommand(). Create a user with dbOwner role on the target database:

use yourdb
db.createUser({
  user: "monitor_slowsql",
  pwd: "123456",
  roles: [{role: "dbOwner", db: "yourdb"}]
})

Tool Deployment

Clone the repository and place it under the web server document root:

git clone https://github.com/hcymysql/mongo_slowquery.git /var/www/html/mongo_slowquery

1. Import MySQL schema

mysql -uroot -p123456 < mongo_slowsql_schema.sql

2. Register monitored MongoDB instance

INSERT INTO mongo_status_info
(ip, tag, user, pwd, port, dbname, threshold_slow_ms)
VALUES
('10.10.159.31','MongoDB Test 1','monitor_slowsql','123456','27017','yourdb',1000);

Adjust fields to match your environment (IP, tag, credentials, port, authentication database, and slow‑query threshold in milliseconds).

3. Configure MySQL connection for the web UI

Edit /var/www/html/mongo_slowquery/conn.php:

$con = mysqli_connect("127.0.0.1","admin","123456","mongo_monitor","3306")
       or die("Database connection error: " . mysqli_error());

4. Schedule periodic collection

Add a cron entry to run the collector every ten minutes:

*/10 * * * * cd /var/www/html/mongo_slowquery && /usr/bin/php check_mongo_slowsql.php > /dev/null 2>&1

5. Access the UI

Open a browser and navigate to:

http://<em>yourIP</em>/mongo_slowquery/mongo_slowquery.php

The dashboard shows aggregated slow‑query statistics for the last 31 days, allows filtering by tag, and provides detailed information (user, client IP, collection size, indexes, and explain plan) for each operation.

Enable Slow Query Logging in MongoDB

Set profiling level to capture operations longer than 1 second: db.setProfilingLevel(1, 1000); Verify the setting:

db.getProfilingStatus();

Repository

Source code: https://github.com/hcymysql/mongo_slowquery