MySQL Proxy and microseconds

By default the included Lua within MySQL proxy (0.8.3) does not include socket, necessary for getting microsecond granularity. To setup you have to install Lua and socket on the OS first:

For CentO5

$ sudo yum install lua lua-socket

For Ubuntu

$ sudo apt-get install lua5.1 liblua5.1-socket2

The following enables use within MySQL Proxy.

cp /usr/share/lua/5.1/socket.lua /path/to/mysqlproxy/lib/mysql-proxy/lua
cp -r /usr/lib64/lua/5.1/socket /path/to/mysqlproxy/lib/mysql-proxy/lua
cp -r /usr/lib64/lua/5.1/mime /path/to/mysqlproxy/lib/mysql-proxy/lua

My lua script can now use syntax similar to:

require 'socket'

function read_query( packet )
  print( string.format("# %s.%3dn%s;n","%X",now),select(2,math.modf(now))*1000 , query))

Announcing the MySQL Plugin for New Relic

Many application developers would know of New Relic. A SaaS performance and monitoring tool targeted towards your web application monitoring including PHP, Ruby, Java, .Net, Python and Node.

With the release today (June 19, 2013) of the New Relic Platform, custom monitoring of data stores including MySQL are now possible. Try it now free. This link will provide you a free standard account (no cost, no billing details necessary), that enables you to perform application monitoring, server monitoring, MySQL instance monitoring and monitoring of many other products via many plugins.

Over the next few posts I will be discussing some of the design decisions I made for this MySQL plugin. New Relic has certain features that lend towards really helping developers monitor and diagnose the application (I have been surprised how it has helped in debugging DB and OS problems directly from PHP code for example). However, often it’s important that knowing the server resources, the database usage is critical to seeing the whole picture, and with the new plugins, New Relic gives developers, system administrators and database administrator some well targeted insights. When building custom dashboards you can see CPU usage, Database usage, and your web application volume, page load time and other metrics on one page.

The MySQL plugin has two pre-requisite requirements. A MySQL server running 5.0 or better, and a Java JRE 1.6 or better. The plugin can work either directly on each MySQL server/instance, and therefore needs the JRE, or it is possible to configure a single server to collect all MySQL statistics and report them to New Relic. There are no specifics that restrict this plugin working for any MySQL variants/forks, infact I specifically designed the plugin to be forward compatible with new version and status variables for example, and support custom recording of any metrics (more on that later).

This is first release of the New Relic Platform and MySQL plugin so I expect a lot of refinements, improvements and suggestions as we move forward. As an integral part of developing the MySQL plugin and using the New Relic Java SDK, there are a number of roadmap items to better serve MySQL and other products that will be coming in future releases. The beta version of the MySQL plugin has been running on production MySQL servers for several months now and working well.

New Relic provides two ways to display data, first by graphs, and second by tables. There is a handy information option in the Server monitoring that is not presently available for the plugins. Graphs work best with multiple data points and constantly changing data and records averages. When looking at the SQL commands for example is great to see the total breakdown, monitoring MySQL replication lag (a single metric), that hopefully is generally zero can appear a little bland. A cool trick is to click on the legend, this toggles the displayed value, and can really help when one value in a graph hogs the metric.

At present the plugin has 4 tabs of display:

  • Overview provides a high level view of total reads v total writes, database connections, network communications and a table of key utilizations (which I will discuss more in another post)
  • Query Analysis shows details of SQL commands, temporary (memory/disk) tables, slow queries, query cache usage, select query types, sort types and table locking.
  • InnoDB Metrics include buffer pool operations, a page breakdown, row operations, log writes, log and data fsyncs, checkpoint age, history list, internal threads and mutexes.
  • Replication shows lag, relay log volume, I/O thread lag, slave errors and master binary log volume (when on a master).

The plugin is written to be extensible via JSON configuration. Those wishing to monitor different variables, or say custom metrics from storage engines like Tokutek can be easily defined, either a key/value set, or single row of metrics. However, the initial version of the dashboards does not allow the customers to modify the present dashboard. Requests are welcome for me to expand the current global dashboards.

In the News

Percona Ireland??

Anybody else noticed that Percona appears to not be a US entity any more?

I observed it today.

$ sudo /usr/bin/innobackupex ...
InnoDB Backup Utility v1.5.1-xtrabackup; Copyright 2003, 2009 Innobase Oy
and Percona Ireland Ltd 2009-2012.  All Rights Reserved.

This software is published under
the GNU GENERAL PUBLIC LICENSE Version 2, June 1991.

In previous versions this did say Percona Inc, 2009-2012.

The footer of the Documentation also states © Copyright Percona Ireland Ltd. 2009-2013.

Percona Live Conference Recommendations

Percona Live MySQL Conference and Expo, April 22-25, 2013

While many attendees are repeat offenders, if 2013 is your first MySQL conference and you are relatively new with MySQL (say < 2 years experience), it can be daunting to determine which of the 8 or more concurrent sessions you should attend during the conference.

Here are my top recommendations that give you a good grounding in the various conference topics and a wealth of experience from known MySQL authorities, on important topics.

  1. A backup today saves you tomorrow by Ben Mildred at Pythian. Losing your data is a terrible experience. Learn what is needed to keep your data safe and you system highly available.
  2. Survey of Percona Toolkit: Command-Line tools for MySQL by Daniel Nichter at Percona. There are a wealth of additional MySQL tools that any resource should be familiar with. These are some of the most popular.
  3. Script It. Make Professional DBA tools out of nothing by Giuseppe Maxia at Continuent. I seasoned expert in the MySQL field, his expertise is invaluable to learn how to use MySQL effeciently. Giuseppe is also the creator of MySQL Sandbox, a huge productivity tool for developers.
  4. Practices for reducing MySQL database size by Yoshinori Matsunobu at Facebook. This is a consulting technique I use for great advantage with clients to improve performance. Yoshinori is also one of the most popular technical speakers at events.
  5. MHA: Getting started and moving past the quirks by Colin Charles at Monty Program. Creating a HA environment is essential for any successful application. MHA is one open source approach that should be considered.
  6. Managing data and data archiving using MySQL 5.6 new features of portable tablespace and exchange partition by Marco tusa at Pythian. If there was one footnote feature in MySQL 5.6 that has a huge benefit, this is the feature. As data continues to grow rapidly in size, archiving is more important.

This year’s conference talks are organized by topic and skill level. This can also help you find talks specific to your needs. Topics include the following:

  • Developing Applications
  • Tools
  • Best Practices for Businesses
  • Database Administration
  • Utilizing Hardware
  • Replication and High Availability Strategies
  • Treads in Architecture and Design
  • New Features

2013 is sure to be a great event, with a lineup of many MySQL product features for the MySQL ecosystem.

When is the error log filename not the right filename

When evaluating a MySQL system one of the first things to look at is the MySQL error log. This is defined by the log[_-]error variable in the MySQL Configuration file. Generally found like:

grep log.error /etc/my.cnf

It is possible to find multiple rows because this could be defined in the [mysqld] and [mysqld_safe] sections. It is also possible it is incorrectly defined twice in any given section.

Immediately I see a problem here, and the following describes why. If you look at this file name, in this case it’s actually found, but the file is empty.

$ ls -l /var/lib/mysql/logs/mysql_error_log
-rw-r----- 1 mysql mysql 0 Feb 19 20:35 /var/lib/mysql/logs/mysql_error_log

An error log should never exist and be empty, because starting the instance producing messages. An error log could be empty because the system does a log rotate (BTW, never rotate the error log, see The correct approach to rolling MySQL logs). However it is empty in this case because MySQL is not writing to the error log filename as defined, because it does not have file extension. NOTE: there is no .log or similar extension. Looking more closely.

$ ls -l /var/lib/mysql/logs/mysql_error_log*
-rw-r----- 1 mysql mysql      0 Feb 19 20:35 /var/lib/mysql/logs/mysql_error_log
-rw-rw---- 1 mysql root  394530 Feb 19 20:35 /var/lib/mysql/logs/mysql_error_log.err

As you can see, MySQL has overwritten your parameter and given the file an extension, as verified by the runtime value.

mysql> show global variables like 'log_error';
| Variable_name | Value                                   |
| log_error     | /var/lib/mysql/logs/mysql_error_log.err |

So, the tip is, always use .log for your MySQL error (and slow query log).

NOTE: Placing the log files in the MySQL datadir (which defaults to /var/lib/mysql) is also a bad idea. A topic for another blog post at a later time.

The MySQL symlink trap

Many users of MySQL install and use the standard directories for MySQL data and binary logs. Generally this is /var/lib/mysql.
As your system grows and you need more disk space on the general OS partition that commonly holds /tmp, /usr and often /home, you create a dedicated partition, for example /mysql. The MySQL data, binary logs etc are then moved to this partition (hopefully in dedicated directories). For example data is placed in /mysql/data.
Often however, a symbolic link (symlink) is used to so MySQL still refers to the data in /var/lib/mysql.

When it comes to removing the symlink and correctly configuring MySQL, you first stop MySQL and correctly defining the datadir my.cnf variable to point to the right location. However, MySQL still keeps the legacy directory information around and this will cause MySQL replication to fail in several ways when you attempt to restart your MySQL instance.

The binary log index, the relay log index, and the relay log info files all contain the legacy path. MySQL does not make it easy to also determine these actual files.

The relay_log_index variable defines the index, but defaults to [relay_log].index when not defined, so with SHOW GLOBAL VARIABLES this may be blank.
log-bin-index is an configurable option, but no matching global variable. It defaults to [log-bin].index.
relay_log_info will contain a value, generally only a file that is relevant to the data directory.

In these situations, your only option to to manually edit these files, specifying the new datadir (or log-bin) path in order to correctly remove symlinks.

The best advice, is to consider the design of your system first, and never place data in default locations if you feel this has to be modified later. Define those dedicated directories before you start using your MySQL instance.

Poor programming practices

When will it stop. These amateur programmers that simply cut/paste code really affect those good programmers in the ecosystem trying to make a decent living. I was reviewing a developed (but incomplete) PHP/MySQL system using a common framework (which in itself is irrelevant for this post).

In one source file there were 12 repetitions of the following code:

    if (!array_key_exists($id,$this->session->userdata['permissions']) OR
	!array_key_exists('id', $this->session->userdata['permissions'][$id]) OR
	!array_key_exists('scope', $this->session->userdata['permissions'][$id]['name'])){
      $this->session->set_flashdata('alert', 'You are not authorized to go there.');

It’s bad enough when code is repeated and not put in a simple re factored function. When it’s repeated 12 times in one file, and OMG over 100 times in the product, that is a recipe for bugs, and high maintenance codes due to extremely poor coding practice.

Carbonite Online Backup is a fraud

Do not listen to the hype or the advertising. Carbonite backup solution is a fraud. I never realized the extent of the failures of the software until I had a problem, which is when you expect and demand commercial software you pay for to work.

Ironically, looking now via Google search for Carbonite restore problems there are plenty of horror stories. And just to add to the experience, the definition of Carbonite in the dictionary is “Explosive”.

Here is the first red flag. You logon to the website, and if you click on “View Files”, or under the “Backup” tab with a nice cloud icon you click “Access Files”, there is no information available. You will receive the error “We are unable to access your files on this computer right now. Please contact Customer Support at for further assistance.”. What is really means is “Until your computer is online and your Carbonite software is working, you have no access to the details of your files that are apparently backed up.”

When you contact customer support, they have no idea what that message means, and after wasting your time (for me in a chat session), a ticket was opened with technical support. The problem is there is no way to track your ticket online, get updates, post information etc. I minimized the window, and now my chat session is closed (most likely by the rather ill informed customer representative). Guess what? There is no information about the ticket number in your account. So I may as not every had that conversation, any evidence of it is now lost.

I was told that until a technical support person could access my computer there was no way I could access my files. WTF? If there is no centralized list or log of my files on your backup solution, and no way to see this, how do I know you ever backed up my files. A fancy progress bar that flashes and says backing up files. Any 2 year old can write that faux display. Red flag number 2. It seems the only way to see my files is to install this software on yet another system to restore files. That is as Google Searching indicates, a likely lesson in extreme frustration

The ultimate cause of the problem was my system crashed, and when it restarted, Cabonite software was in this stuck state of “Registering”. I was told to just re-install the software, that’s not an answer in my books. That is red flag number 3.

I have definitely removed my credit card from their site to stop any automatic renewal of this crap software.

For the record, my home office backup solution includes important files on a Drobo. Backups of Documents to DropBox and then sync’d to another system. Backup of all files on several machines to a central external USB, and then regular backups of that which are taken offsite.

You cannot be too careful with important things like photos. Unfortunately this solution lacks a central catalog, and versioning of files (I.e. I overwrote an important presentation and did not realize for a few months, when I looked at all my backups of this, 3 or 4 copies, they were all the overwritten file, not the original. It took about a day to actually find a copied version, not a backed up version)

Upgrading to MySQL 5.5 on Ubuntu 10.04 LTS

Ubuntu does not provide an apt-get repository package for MySQL 5.5 on this older OS, however this is still a widely used long term support version. The following steps will upgrade an existing MySQL 5.1 apt-get version to a standard MySQL 5.5 binary.

Step 1. Remove existing MySQL 5.1 retaining data and configuration

sudo su -
service mysql stop
cp -r /etc/mysql /etc/mysql.51
cp -r /var/lib/mysql /var/lib/mysql.51
which mysqld
dpkg -P mysql-server mysql-server-5.1 mysql-server-core-5.1
which mysqld
which mysql
dpkg -P mysql-client-5.1 mysql-client-core-5.1
which mysql
dpkg -P libdbd-mysql-perl libmysqlclient16 mysql-common
# This will not remove /etc/mysql if any other files are in the directory
dpkg -l | grep mysql
[ -d /etc/mysql ] && mv /etc/mysql /etc/mysql.uninstall
cp -r /etc/mysql.51 /etc/mysql

2. Prepare configuration and required directories.

sudo su -
grep basedir ${MYCNF}
sed -ie "s/^basedir.*$/basedir=/opt/mysql/" ${MYCNF}
grep basedir ${MYCNF}
sed -ie "/^[mysqld_safe]/a
skip-syslog" ${MYCNF}
chown -R mysql:mysql /var/lib/mysql
mkdir -p /var/run/mysqld
sudo chown mysql:mysql /var/run/mysqld

Install MySQL 5.5

sudo su -
mkdir -p /opt
cd /opt
# Install MySQL 5.5 Binaries
apt-get install -y libaio-dev  # New 5.5 dependency
tar xvfz mysql*.tar.gz
ln -s mysql-5.5.28-linux2.6-x86_64 /opt/mysql
echo "export MYSQL_HOME=/opt/mysql
export PATH=$MYSQL_HOME/bin:$PATH" > /etc/profile.d/
chmod +x /etc/profile.d/
. /etc/profile.d/
echo $PATH
which mysql

4. Upgrade and verify MySQL Instance

su - mysql
bin/mysqld_safe --skip-syslog &
tail /var/log/mysql/error.log
# There will be some expected ERRORS in error log
bin/mysql_upgrade -uroot
bin/mysqladmin -uroot  shutdown
bin/mysqld_safe --skip-syslog &
tail -100 /var/log/mysql/error.log
mysql -uroot -e "SELECT VERSION();"
bin/mysqladmin -uroot  shutdown

5. Setup MySQL for system use

# As Root
sudo su -
cp /opt/mysql/support-files/mysql.server ${INIT}
sed -ie "s/^basedir=$/basedir=/opt/mysql/;s/^datadir=$/datadir=/var/lib/mysql/" ${INIT}
${INIT} start
mysql -uroot -e "SELECT VERSION();"
${INIT} stop

The heavy handed LinkedIn approach to your contacts

I recently wanted to add two individuals to my list of professional contacts at LinkedIn. I was extremely disappointed at the modified user interface (UI) experience that made it difficult to do so. In the past, you just entered a list of emails.

Many companies these days pressure you into opening up your entire network of contacts for their benefits of knowing your social graph. This is unacceptable.

You have to go thru the following complexity just to send an email request for connection in LinkedIn now.

  • Add Connections
  • Select any email (last button of options)
  • Click Invite by individual email (hidden at bottom of page)

Not a cool new feature for Master_Host

I was surprised to find on a customer MySQL server this new syntax for Master_host in SHOW SLAVE STATUS.

*************************** 1. row ***************************
               Slave_IO_State: Connecting to master
                  Master_Host: or 10.XXX.XX.XXX
                  Master_User: repl
                  Master_Port: 3306
                Connect_Retry: 60
              Master_Log_File: db1-354215-bin-log.000005
          Read_Master_Log_Pos: 1624
               Relay_Log_File: db2-354214-relay-log.000001

Is this a fancy new Percona Server feature? No. It’s operator error.

We read a little further to find.

             Slave_IO_Running: Connecting

                Last_IO_Errno: 2005
                Last_IO_Error: error connecting to master '[email protected] or 10.XXX.XX.XXX' - retry-time: 60  retries: 86400

How can this be created.
Using MySQL MHA, you get the following message in the output of commands to manage replication.

Thu Jan  3 17:06:40 2013 - [info]  All other slaves should start replication from here.
Statement should be: CHANGE MASTER TO MASTER_HOST=' or 10.XXX.XX.XXX', MASTER_PORT=3306,
 MASTER_LOG_FILE='db1-354215-bin-log.000005', MASTER_LOG_POS=1624, MASTER_USER='repl',

Needless to say, this syntax was taken literately, and MySQL did not complain.

I would suggest here that while MySQL does not do any validation on the value of the MASTER_HOST value in the CHANGE MASTER TO command to ensure it is resolvable it should at least do some validation to ensure the value is either a DNS entry or an IPV4,IPV6 value, that is space ‘ ‘, is not a valid character in these situations.

Installing MySQL MHA with Percona Server

MySQL MHA by Oracle ACE Director Yoshinori Matsunobu is an excellent open source tool to help in providing HA with native MySQL replication. The installation however is dependent on some Perl packages and to the untrained eye this may be an issue if you are using Percona Server as your choice of MySQL implementation.

The MHA Node page requires the perl-DBD-MySQL package to be installed. The installation on RedHat/CentOS/Oracle Linux look like this:

$ sudo yum install perl-DBD-MySQL
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base:
* extras:
* updates:
Setting up Install Process
Resolving Dependencies
--> Running transaction check
---> Package perl-DBD-MySQL.x86_64 0:3.0007-2.el5 set to be updated
--> Finished Dependency Resolution

However, recently with a client the following error “mysql conflicts with Percona-Server” occurred.

$ sudo yum install perl-DBD-MySQL
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base:
* extras:
* updates:
Setting up Install Process
Resolving Dependencies
--> Running transaction check
---> Package perl-DBD-MySQL.x86_64 0:3.0007-2.el5 set to be updated
--> Processing Dependency: for package: perl-DBD-MySQL
--> Processing Dependency: for package: perl-DBD-MySQL
--> Running transaction check
---> Package mysql.x86_64 0:5.0.95-1.el5_7.1 set to be updated
--> Processing Conflict: mysql conflicts MySQL
--> Finished Dependency Resolution
mysql-5.0.95-1.el5_7.1.x86_64 from updates has depsolving problems
 --> mysql conflicts with Percona-Server-client-55
Error: mysql conflicts with Percona-Server-client-55

This is actually an easy problem to solve, but if you are unfamiliar with all the various MySQL packages for RPM’s, and you installed Percona Server with the standard documented commands you may get lost. What is necessary is the installation of the compatibility libraries, as simple as:

$ sudo yum install -y Percona-Server-shared-compat

SQLSTATE[HY000]: General error: 2006 MySQL server has gone away

This would have to be one of the most common MySQL error messages that is misleading to the end user developer. The MySQL Manual page confirms the broad range of possible conditions, but offers little to a PHP developer that does not speak MySQL Geek. I am commonly asked to help solve this issue from a developer.

The problem is that there are several conditions that can cause this error, and a more meaningful explanation to the end user would help in addressing the issue. In general terms, this actually means “Your SQL statement has failed because the connection to the database has been disconnected because of ???”.

Here are a few common situations and how to check for what “???” is.

1. Your MySQL server really did go away.

We can easily check this by looking at the server uptime and the server error log.

$ mysql -uroot -p -e "show global status like 'uptime';"
| Variable_name | Value |
| Uptime        | 68928 |
1 row in set (0.04 sec)
$ tail /var/log/mysql/error.log
130101 22:22:30 InnoDB: Initializing buffer pool, size = 256.0M
130101 22:22:30 InnoDB: Completed initialization of buffer pool
130101 22:22:30 InnoDB: highest supported file format is Barracuda.
130101 22:22:30 InnoDB: 1.1.8 started; log sequence number 63444325509
130101 22:22:30 [Note] Server hostname (bind-address): ''; port: 3306
130101 22:22:30 [Note]   - '' resolves to '';
130101 22:22:30 [Note] Server socket created on IP: ''.
130101 22:22:30 [Note] Event Scheduler: Loaded 0 events
130101 22:22:30 [Note] /usr/sbin/mysqld: ready for connections.
Version: '5.5.28-cll'  socket: '/var/lib/mysql/mysql.sock'  port: 3306  MySQL Community Server (GPL)

In both these cases, the server has been up some time, and there are zero error messages to indicate problems.

If the MySQL server did go away, was it shutdown or did it crash? The MySQL error log will provide the answers. Generally the mysql daemon (mysqld) will be restarted by the mysqld_safe wrapper process.

2. The connection timed out

$ mysql -uroot -p -e "show global variables like '%timeout';"
| Variable_name              | Value    |
| connect_timeout            | 30       |
| delayed_insert_timeout     | 300      |
| innodb_lock_wait_timeout   | 50       |
| innodb_rollback_on_timeout | OFF      |
| interactive_timeout        | 28800    |
| lock_wait_timeout          | 31536000 |
| net_read_timeout           | 30       |
| net_write_timeout          | 60       |
| slave_net_timeout          | 3600     |
| wait_timeout               | 28800    |

These values are relatively sane MySQL defaults. If however you have very short timeouts, you may get this error. Here is just one example.

mysql> SET SESSION wait_timeout=5;

## Wait 10 seconds

mysql> SELECT NOW();
ERROR 2006 (HY000): MySQL server has gone away
No connection. Trying to reconnect...
Connection id:    132361
Current database: *** NONE ***

| NOW()               |
| 2013-01-02 11:31:15 |
1 row in set (0.00 sec)

3. Your SQL statement was killed

Some systems will proactively kill SQL statements that have been running too long. You can easily check if this may be happening proactively by looking at how many KILL statements have been executed.

$ mysql -uroot -p -e "show global status like 'com_kill'"
| Variable_name | Value |
| Com_kill      | 0     |

Not killed this time.

4. Your SQL statement was too large.

A little harder to test and verify, but MySQL uses a maximum packet site for communications between the server and the client. If this includes large fields (for example BLOB columns), you may be getting a termination of your SQL statement due to size.

By default this is relatively small.

mysql> show global variables like 'max_allowed_packet';
| Variable_name      | Value   |
| max_allowed_packet | 1048576 |
1 row in set (0.00 sec)

You can increase, for example to 16M with:

mysql> set global max_allowed_packet=1024*1024*16;
mysql> show global variables like 'max_allowed_packet';
| Variable_name      | Value    |
| max_allowed_packet | 16777216 |
1 row in set (0.00 sec)

The good news, is this was the cause for the customer today, and now no more errors!

Be sure to keep this value during MySQL restarts.

max_allowed_packet = 16M