Databases, Systems & Networks

Running MySQL at optimal settings for specific resources helps handle larger server loads and prevents server slowdown. Generally, after tuning Apache to handle larger loads, it is beneficial to tune MySQL to additional connections.

Database tuning is an expansive topic, and this guide covers only the basics of editing your MySQL configuration. Large MySQL databases can require a considerable amount of memory. For this reason, we recommend using a high memory Linode for such setups.

The steps in this guide require root privileges. Be sure to run the steps below as root or with the sudo prefix. For more information on privileges see our Users and Groups guide.

Tools That Can Help Optimize MySQL

In order to determine if your MySQL database needs to be reconfigured, it is best to look at how your resources are performing now. This can be done with the top command or with the Linode Longview service. At the very least, you should familiarize yourself with the RAM and CPU usage of your server, which can be discovered with these commands:

1
2

echo [PID]  [MEM]  [PATH] &&  ps aux | awk '{print $2, $4, $11}' | sort -k2rn | head -n 20
ps -eo pcpu,pid,user,args | sort -k 1 -r | head -20

MySQLTuner

The MySQLTuner script assesses your MySQL installation, and then outputs suggestions for increasing your server’s performance and stability.

1. Download and run MySQLTuner:

   1	curl -L http://mysqltuner.pl/ | perl

2. It outputs your results:

   1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

   >> MySQLTuner 1.4.0 - Major Hayden <major@mhtx.net> >> Bug reports, feature requests, and downloads at http://mysqltuner.com/ >> Run with '--help' for additional options and output filtering Please enter your MySQL administrative login: root Please enter your MySQL administrative password: [OK] Currently running supported MySQL version 5.5.41-0+wheezy1 [OK] Operating on 64-bit architecture -------- Storage Engine Statistics ------------------------------------------- [--] Status: +ARCHIVE +BLACKHOLE +CSV -FEDERATED +InnoDB +MRG_MYISAM [--] Data in InnoDB tables: 1M (Tables: 11) [--] Data in PERFORMANCE_SCHEMA tables: 0B (Tables: 17) [!!] Total fragmented tables: 11 -------- Security Recommendations ------------------------------------------- [OK] All database users have passwords assigned -------- Performance Metrics ------------------------------------------------- [--] Up for: 47s (113 q [2.404 qps], 42 conn, TX: 19K, RX: 7K) [--] Reads / Writes: 100% / 0% [--] Total buffers: 192.0M global + 2.7M per thread (151 max threads) [OK] Maximum possible memory usage: 597.8M (60% of installed RAM) [OK] Slow queries: 0% (0/113) [OK] Highest usage of available connections: 0% (1/151) [OK] Key buffer size / total MyISAM indexes: 16.0M/99.0K [!!] Query cache efficiency: 0.0% (0 cached / 71 selects) [OK] Query cache prunes per day: 0 [OK] Temporary tables created on disk: 25% (54 on disk / 213 total) [OK] Thread cache hit rate: 97% (1 created / 42 connections) [OK] Table cache hit rate: 24% (52 open / 215 opened) [OK] Open file limit used: 4% (48/1K) [OK] Table locks acquired immediately: 100% (62 immediate / 62 locks) [OK] InnoDB buffer pool / data size: 128.0M/1.2M [OK] InnoDB log waits: 0 -------- Recommendations ----------------------------------------------------- General recommendations: Run OPTIMIZE TABLE to defragment tables for better performance Enable the slow query log to troubleshoot bad queries Variables to adjust: query_cache_limit (> 1M, or use smaller result sets)

   MySQLTuner offers suggestions regarding how to better the database’s performance. If you are wary about updating your database on your own, following MySQLTuner’s suggestions is one of the safer ways to improve your database performance.

Tuning MySQL

When altering the MySQL configuration, be alert to the changes and how they affect your database. Even when following the instructions of programs such as MySQLTuner, it is best to have some understanding of the process.

The file you are changing is located at /etc/mysql/my.cnf.

Prior to updating the MySQL configuration, create a backup of the my.cnf file:

1	cp /etc/mysql/my.cnf ~/my.cnf.backup

Best practice suggests that you make small changes, one at a time, and then monitor the server after each change. You should restart MySQL after each change:

• For systems without systemd:

  1	systemctl restart mysqld

• For distributions which don’t use systemd:

  1	service mysql restart

When changing values in the my.cnf file, be sure that the line you are changing hasn’t been commented out with the pound (#) prefix.

key_buffer

Changing the key_buffer allocates more memory to MySQL, which can substantially speed up your databases, assuming you have the memory free. The key_buffer size should generally take up no more than 25 percent of the system memory when using the MyISAM table engine, and up to 70 percent for InnoDB. If the value is set too high, resources are wasted.

According to MySQL’s documentation, for servers with 256MB (or more) of RAM with many tables, a setting of 64M is recommended. Servers with 128MB of RAM and fewer tables can be set to 16M, the default value. Websites with even fewer resources and tables can have this value set lower.

max_allowed_packet

This parameter lets you set the maximum size of a sendable packet. A packet is a single SQL state, a single row being sent to a client, or a log being sent from a master to a slave. If you know that your MySQL server is going to be processing large packets, it is best to increase this to the size of your largest packet. Should this value be set too small, you would receive an error in your error log.

thread_stack

This value contains the stack size for each thread. MySQL considers the default value of the thread_stack variable sufficient for normal use; however, should an error relating to the thread_stack be logged, this can be increased.

thread_cache_size

If thread_cache_size is “turned off” (set to 0), then any new connection being made needs a new thread created for it. When the connections disengage the thread is destroyed. Otherwise, this value sets the number of unused threads to store in a cache until they need to be used for a connection. Generally this setting has little affect on performance, unless you are receiving hundreds of connections per minute, at which time this value should be increased so the majority of connections can be made on cached threads.

max_connections

This parameter sets the maximum amount of concurrent connections. It is best to consider the maximum amount of connections you have had in the past before setting this number, so you’ll have a buffer between that upper number and the max_connections value. Note, this does not indicate the maximum amount of users on your website at one time; rather it shows the maximum amount of users making requests concurrently.

table_cache

This value should be kept higher than your open_tables value. To determine this value use:

1

SHOW STATUS LIKE 'open%';

As with all relational databases, MySQL can prove to be a complicated beast, one that can crawl to a halt at a moment’s notice, leaving your applications in the lurch and your business on the line.

The truth is, common mistakes underlie most MySQL performance problems. To ensure your MySQL server hums along at top speed, providing stable and consistent performance, it is important to eliminate these mistakes, which are often obscured by some subtlety in your workload or a configuration trap.

Luckily, many MySQL performance issues turn out to have similar solutions, making troubleshooting and tuning MySQL a manageable task.

Here are 10 tips for getting great performance out of MySQL.

MySQL performance tip No. 1: Profile your workload

The best way to understand how your server spends its time is to profile the server’s workload. By profiling your workload, you can expose the most expensive queries for further tuning. Here, time is the most important metric because when you issue a query against the server, you care very little about anything except how quickly it completes.

The best way to profile your workload is with a tool such as MySQL Enterprise Monitor’s query analyzer or the pt-query-digest from the Percona Toolkit. These tools capture queries the server executes and return a table of tasks sorted by decreasing order of response time, instantly bubbling up the most expensive and time-consuming tasks to the top so that you can see where to focus your efforts.

Workload-profiling tools group similar queries together, allowing you to see the queries that are slow, as well as the queries that are fast but executed many times.

Lire la suite…

Source: digitalocean

Introduction

MySQL comes with a handy little diagnostic tool called mysqlslap that’s been around since version 5.1.4. It’s a benchmarking tool that can help DBAs and developers load test their database servers.

mysqlslap can emulate a large number of client connections hitting the database server at the same time. The load testing parameters are fully configurable and the results from different test runs can be used to fine-tune database design or hardware resources.

In this tutorial we will learn how to use mysqlslap to load test a MySQL database with some basic queries and see how benchmarking can help us fine-tune those queries. After some basic demonstrations, we will run through a fairly realistic test scenario where we create a copy of an existing database for testing, glean queries from a log, and run the test from a script.

The commands, packages, and files shown in this tutorial were tested on CentOS 7. The concepts remain the same for other distributions. Lire la suite…

One of the interesting new tools in MySQL 5.1.4 is mysqlslap, a load emulator that lets you see how well a particular query set or table engine performs under high-load conditions.

A query that consumes too many database resources may be the result of designing tables incorrectly, choosing the wrong table type, or creating an inefficient query. When a query eats up a lot of database resources, it can negatively affect other application components. By using mysqlslap to stress test a server in a non-public environment, you will discover these errors sooner, allowing you to you avoid a database meltdown once your application goes live.

This tutorial shows how you can use mysqlslap to run stress tests involving multiple clients, custom queries, different table engines, and much more. Lire la suite…