Pablowe

Nagios Checks For MMM

I’ve written some new Nagios checks for MMM (MMM on Google Code … MMM on Launchpad). check_mmm is a part of http://code.google.com/p/check-mysql-all/, and is meant to be called locally on the MMM Monitor server (usually via NRPE). Feedback is welcome, usage is as follows:

Usage:
     check_mmm --cluster C# 

     Options:
       --cluster=    The MMM Cluster to check
       -c, --critical=
    The level at which a critical alarm is raised.
       -h, --help                Display this message and exit
       -v, --verbose             Increase verbosity level
       -V, --version             Display version information and exit
       -w, --warning             The level at which a warning is raised.

     Defaults are:

     ATTRIBUTE                  VALUE
     -------------------------- ------------------
     cluster                    No default value
     critical                   HARD_OFFLINE,REPLICATION_FAIL
     help                       FALSE
     verbose                    1 (out of 3)
     version                    FALSE
     warning                    ADMIN_OFFLINE,AWAITING_RECOVERY,REPLICATION_DELAY

Nagios MySQL Plug-Ins

There currently exist many plugins for MySQL to use with Nagios. Many of them, however, are not version-independent, leaving organizations that use multiple versions of MySQL to either install multiple plugins or not monitor specific versions of MySQL. As such, I’ve compiled what I consider to be the most useful checks into a single plugin: check_mysql

Usage:
     check_mysql check_name [options]

     Options:
       --args|a     Optional arguments.  Comma-separated.  Check-specific.
       --critical|c The level at which a critical alarm is raised.  Check-specific.
       --database   The database to use (defaults to mysql)
       --help|?     Display this message and exit
       --hostname|H     The target MySQL server host (defaults to localhost)
       --password|p The password of the MySQL user
       --port       The port MySQL is listening on (defaults to 3306)
       --user|u     The MySQL user used to connect
       --version|V  Display version information and exit
       --warning|w  The level at which a warning is raised.  Check-specific.

     defaults are:

     ATTRIBUTE                  VALUE
     -------------------------- ------------------
     args                       No default value
     critical                   Check-specific
     database                   mysql
     help                       FALSE
     host                       localhost
     password                   No default value
     port                       3306
     timeout                    10 seconds
     user                       No default value
     verbose                    1 (out of 3)
     version                    FALSE
     warning                    Check-specific

Current Checks Supported:

* connect – Check to see whether or not one can connect to MySQL (USAGE)
* repl_io – Check to see whether on not the IO Replication thread is running (REPLICATION CLIENT)
* repl_sql – Check to see whether or not the SQL Replication thread is running (REPLICATION CLIENT)
* repl_sbm – Check how many seconds behind the master the slave is (REPLICATION CLIENT)
* mysql_query – Run a given query, test if it executes properly (SELECT)
* connections – Test if the percentage of used connections is over a given threshold (PROCESS)

I am open to requests for additional checks etc.

OmniSQL 0.0.7 Released

OmniSQL (a command line tool for DBAs needing to issue ad-hoc queries against sharded data) version 0.0.7 is officially released.

Instead of logging in separately to multiple databases to issue the same query, groups of databases can be specified in a configuration file and queries will be automatically issued against all targeted MySQL instances.

Let me know of any bugs found or features you would like to see in upcoming releases!

Download at http://code.google.com/p/omnisql/.

CHANGE LOG
- Fixed Bug #3: Script failure when only one group is defined
- Fixed Bug #4: Too Slow (now it forks)
- Fixed Bug #5: Host-Specific parameters don't work
- INCOMPATIBLE CHANGE: Changed format of config file
- INCOMPATIBLE CHANGE: Use XML::SimpleObject instead of XML::Simple
- Added group summary (Issue #6)

Managing MySQL Configuration Files

It is good practice to manage changes to MySQL configuration files (/etc/my.cnf) by using a version control system. I usually use a home-brewed (not brewed by me!) svn+cfengine application to propagate my.cnf (and other configuration file) changes to defined classes of machines (classes are based on application role, replication role, etc).

When managing hundreds of different database servers with dozens of roles, templating my.cnf becomes a necessity! If I have to change a variable for a given class, that can mean editing a dozen my.cnf files on as many servers. Not a productive use of time!!! Usually, the only parameters that differ between database hosts in the same class are the replication options. As such, I find it useful to have a base template for each class of machine and use the !include startup option to specify host-specific startup options.

Here is a sample of what I would have as a base my.cnf for billing-class MySQL Replication Slaves:

And then I would store host-specific replication options in a smaller file:

To re-iterate, templating MySQL configuration files can help ensure consistency between servers of the same class as well as making it easier to push changes to all relevant hosts instead of doing it individually.