在MYSQL查询中使用WHERE和ORDER BY条件时结果非常慢

Question

I am facing issue of very slow result.

I am sharing table structure as and results also.

if you have any suggestion please update soon if possible for you.

================================================================= Table Structure -- https://www.phpmyadmin.net/ -- -- Host: localhost:3306 -- Generation Time: Mar 25, 2019 at 11:48 AM -- Server version: 5.5.61-cll -- PHP Version: 7.2.7

SET SQL_MODE = "NO_AUTO_VALUE_ON_ZERO";
SET AUTOCOMMIT = 0;
START TRANSACTION;
SET time_zone = "+00:00";

--
-- Database: `energe3c_lms`
--

-- --------------------------------------------------------

--
-- Table structure for table `user_material`
--

CREATE TABLE `user_material` (
  `id` int(11) NOT NULL,
  `user_course_id` int(11) NOT NULL,
  `user_id` int(11) NOT NULL,
  `material_id` int(11) NOT NULL,
  `attempt` int(11) NOT NULL,
  `another_attempt` tinyint(1) DEFAULT '0',
  `status` varchar(255) DEFAULT NULL,
  `complete` tinyint(1) DEFAULT NULL,
  `percent` float DEFAULT '0',
  `time` varchar(255) DEFAULT NULL,
  `marking_time` varchar(255) DEFAULT NULL,
  `marked_by` int(11) DEFAULT NULL,
  `feedback` text,
  `submitted_date` datetime DEFAULT NULL,
  `marking_date` datetime DEFAULT NULL,
  `created` datetime DEFAULT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

--
-- Indexes for dumped tables
--

--
-- Indexes for table `user_material`
--
ALTER TABLE `user_material`
  ADD PRIMARY KEY (`id`),
  ADD KEY `user_material-user` (`user_id`),
  ADD KEY `user_material-material` (`material_id`),
  ADD KEY `user_material-marking-user` (`marked_by`),
  ADD KEY `user_course-user_material` (`user_course_id`);

--
-- AUTO_INCREMENT for dumped tables
--

--
-- AUTO_INCREMENT for table `user_material`
--
ALTER TABLE `user_material`
  MODIFY `id` int(11) NOT NULL AUTO_INCREMENT;

--
-- Constraints for dumped tables
--

--
-- Constraints for table `user_material`
--
ALTER TABLE `user_material`
  ADD CONSTRAINT `user_course-user_material` FOREIGN KEY     (`user_course_id`) REFERENCES `user_course` (`id`) ON DELETE CASCADE ON     UPDATE CASCADE,
  ADD CONSTRAINT `user_material-marking-user` FOREIGN KEY (`marked_by`) REFERENCES `user` (`id`) ON DELETE SET NULL ON UPDATE CASCADE,
  ADD CONSTRAINT `user_material-material` FOREIGN KEY (`material_id`) REFERENCES `material` (`id`) ON DELETE CASCADE ON UPDATE CASCADE,
  ADD CONSTRAINT `user_material-user` FOREIGN KEY (`user_id`)     REFERENCES `user` (`id`) ON DELETE CASCADE ON UPDATE CASCADE;

COMMIT;

Showing rows 0 - 24 (25586268 total, Query took 0.0007 seconds.)

SELECT * FROM user_material

=================================================================

Showing rows 0 - 24 (77 total, Query took 22.9434 seconds.)

SELECT * FROM user_material where status='submitted'

=================================================================

Showing rows 0 - 24 (34026300 total, Query took 24.4978 seconds.) [submitted_date: ... - ...]

SELECT * FROM user_material ORDER BY submitted_date ASC

=================================================================

If You have any suggestion, please update me.

I am using this query in yii2 framework

Answer 1

For this query:

SELECT *
FROM user_material 
WHERE status = 'submitted';

You want an index on user_material(status) .

For this query:

SELECT *
FROM user_material
ORDER BY submitted_date ASC

You want an index on user_material(submitted_date) .

In both these cases, additional columns can be in the index after the one specified as the first column. I do not see these indexes in your schema.

Answer 2

Adding to the recommendation from @Gordon Linoff, you should consider normalizing your table further, to make the SQL more efficient. Use an very short code or int for status_id and store a unique list of status codes and descriptions in a lookup table. Then simply join the lookup table in your SQL. See this article for a detailed explanation on normalization.

I also recommend either changing the column name or the data type for the time column, because it doesn't make sense. What time value requires 255 characters?

If you are concerned about performance, then the first thing to focus your attention on is removing unnecessary bytes, sorts and filters from your query. This approach deals with most performance issues. How you join tables is important as well. As the data volume and complexity increases, the tuning effort becomes increasingly more complex as well.

Getting to know EXPLAIN will help your effort. Refer to the EXPLAIN documentation for more information.

-- Create your lookup table first
CREATE TABLE `material_status` (
  `id` int(11) NOT NULL,
  `status` varchar(255) DEFAULT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

ALTER TABLE `material_status`
  ADD PRIMARY KEY (`id`),

ALTER TABLE `material_status`
  MODIFY `id` int(11) NOT NULL AUTO_INCREMENT;

-- Then create your user_material table
CREATE TABLE `user_material` (
  `id` int(11) NOT NULL,
  `user_course_id` int(11) NOT NULL,
  `user_id` int(11) NOT NULL,
  `material_id` int(11) NOT NULL,
  `attempt` int(11) NOT NULL,
  `another_attempt` tinyint(1) DEFAULT '0',
  `status_id` int(11) DEFAULT NULL,
  `complete` tinyint(1) DEFAULT NULL,
  `percent` float DEFAULT '0',
  `time` varchar(255) DEFAULT NULL,
  `marking_time` varchar(255) DEFAULT NULL,
  `marked_by` int(11) DEFAULT NULL,
  `feedback` text,
  `submitted_date` datetime DEFAULT NULL,
  `marking_date` datetime DEFAULT NULL,
  `created` datetime DEFAULT NULL
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

-- Create your extra index here
CREATE INDEX ix-submitted_date ON user_material(submitted_date);

-- Add the status to the keys and constraints
ALTER TABLE `user_material`
  ADD PRIMARY KEY (`id`),
  ADD KEY `user_material-user` (`user_id`),
  ADD KEY `user_material-material` (`material_id`),
  ADD KEY `user_material-marking-user` (`marked_by`),
  ADD KEY `user_course-user_material` (`user_course_id`),
  ADD KEY `user_material-status` (`status_id`);

ALTER TABLE `user_material`
  MODIFY `id` int(11) NOT NULL AUTO_INCREMENT;

ALTER TABLE `user_material`
  ADD CONSTRAINT `user_course-user_material` FOREIGN KEY (`user_course_id`) REFERENCES `user_course` (`id`) ON DELETE CASCADE ON UPDATE CASCADE,
  ADD CONSTRAINT `user_material-marking-user` FOREIGN KEY (`marked_by`) REFERENCES `user` (`id`) ON DELETE SET NULL ON UPDATE CASCADE,
  ADD CONSTRAINT `user_material-material` FOREIGN KEY (`material_id`) REFERENCES `material` (`id`) ON DELETE CASCADE ON UPDATE CASCADE,
  ADD CONSTRAINT `user_material-user` FOREIGN KEY (`user_id`)     REFERENCES `user` (`id`) ON DELETE CASCADE ON UPDATE CASCADE,
  ADD CONSTRAINT `user_material-status` FOREIGN KEY (`status_id`) REFERENCES `material_status` (`id`) ON DELETE CASCADE ON UPDATE CASCADE;

The queries will now look like this:

SELECT
um.id,
um.user_course_id,
um.user_id,
um.material_id,
um.attempt,
um.another_attempt,
ms.status,
um.complete,
um.percent,
um.time,
um.marking_time,
um.marked_by,
um.feedback,
um.submitted_date,
um.marking_date,
um.created
FROM user_material um
INNER JOIN material_status ms on ms.id = um.status_id
WHERE ms.status = 'submitted';

SELECT
um.id,
um.user_course_id,
um.user_id,
um.material_id,
um.attempt,
um.another_attempt,
ms.status,
um.complete,
um.percent,
um.time,
um.marking_time,
um.marked_by,
um.feedback,
um.submitted_date,
um.marking_date,
um.created
FROM user_material um
INNER JOIN material_status ms on ms.id = um.status_id
ORDER BY um.submitted_date ASC;

Answer 3

Please try this.

SELECT
um.id,
um.user_course_id,
um.user_id,
um.material_id,
um.attempt,
um.another_attempt,
ms.status,
um.complete,
um.percent,
um.time,
um.marking_time,
um.marked_by,
um.feedback,
um.submitted_date,
um.marking_date,
um.created
FROM user_material um
INNER JOIN material_status ms on ms.id = um.status_id
AND ms.status = 'submitted'; 

And also create indexing on where and order field.