Liip Monitor Bundle

Installs: 6 875 213

Dependents: 17

Suggesters: 4

Security: 0

Stars: 466

Watchers: 64

Forks: 102

Open Issues: 8


2.23.0 2024-03-05 16:29 UTC


CI Status Scrutinizer Code Quality

This bundle provides a way to run a series of application related health checks. Health checks in the scope of this bundle go beyond simple actions like performing a ping to a server to see if it's alive. For example a Memcache server can be alive and not displaying any errors in your Nagios but you might not be able to access it from your PHP application. Each health check should then implement some application logic that you want to make sure always works. Another usage can be testing for specific requirements, like availability of PHP extensions.

Another design goal of the bundle was to be able to perform the checks using the same configuration and environment that your application is using. In that way you can make sure that if the health check runs successfully then your app should work too.

So each health check will be a class that will implement the CheckInterface::check method which must return a CheckResult object. What happens inside that method is up to the check developer.

Health checks are defined as Symfony services and they have to be tagged as liip_monitor.check in order to be picked up by the health check runner. This gives a lot of flexibility to application and bundle developers when they want to add their own checks.

Checks are run via the command line using a Symfony command or via a REST api that delivers the results in JSON format.

Here's the web interface:

Web Interface


Install with composer:

composer require liip/monitor-bundle

Then register the bundle in the AppKernel.php file:

public function registerBundles()
    $bundles = array(
        // ...
        new Liip\MonitorBundle\LiipMonitorBundle(),
        // ...

    return $bundles;

If you want to enable the REST API provided by the bundle then add the following to your routing.yml:

    resource: "@LiipMonitorBundle/Resources/config/routing.xml"
    prefix: /monitor/health

Then, enable the controller in your configuration:

    enable_controller: true

And finally don't forget to install the bundle assets into your web root:

./app/console assets:install web --symlink --relative

Enabling built-in health checks

To enable built-in health checks, add them to your config.yml

        php_extensions: [apc, xdebug]

Adding Health Checks

See Writing Custom Checks for instructions on creating a custom check.

Once you implemented the class then it's time to register the check service with our service container:

        class: Acme\HelloBundle\Check\PhpExtensionsCheck
            - [ xhprof, apc, memcache ]
            - { name: liip_monitor.check, alias: php_extensions }

The important bit there is to remember to tag your services with the liip_monitor.check tag. By doing that the check runner will be able to find your checks. Keep in mind that checks can reside either in your bundles or in your app specific code. The location doesn't matter as long as the service is properly tagged. The alias is optional and will then simply define the id used when running health checks individually, otherwise the full service id must be used in this case.

If your app's service definition is using autoconfigure to discover services then classes which implement Laminas\Diagnostics\Check\CheckInterface will be tagged automatically.

Available Built-in Health Checks

See "Full Default Config" below for a list of all built-in checks and their configuration.

Running Checks

There are two ways of running the health checks: by using the CLI or by using the REST API provided by the bundle. Let's see what commands we have available for the CLI:

List Checks

$ ./app/console monitor:list


Run All the Checks

$ ./app/console monitor:health

Jackrabbit Health Check: OK
Redis Health Check: OK
Memcache Health Check: KO - No configuration set for session.save_path
PHP Extensions Health Check: OK

Run Individual Checks

To run an individual check you need to provide the check id to the health command:

$ ./app/console monitor:health monitor.check.php_extensions

PHP Extensions Health Check: OK

Run health checks as composer post-install/update scripts

To run health checks as a composer post-install or post-update script, simply add the Liip\\MonitorBundle\\Composer\\ScriptHandler::checkHealth ScriptHandler to the post-install-cmd / post-update-cmd command sections of your composer.json:

"scripts": {
    "post-install-cmd": [
    "post-update-cmd": [

Adding Additional Reporters

There are two default reporters: ArrayReporter for the REST API and ConsoleReporter for the CLI command. You can add additional reporters to be used by either of these.

First, define an additional reporter service and tag it with liip_monitor.additional_reporter:

    class: My\Reporter
        - { name: liip_monitor.additional_reporter, alias: my_reporter }

To run additional reporters with the CLI, add --reporter=... options for each one:

    ./app/console monitor:health --reporter=my_reporter

To run this reporter with the REST API, add a reporters query parameter:


You can list available reporters with:

bin/console monitor:list --reporters

Grouping Checks

It is possible to group the health checks for different environments (e.g. application server, cron runner, ...). If not specified differently, all health checks belong to the default group.

Define groups for build-in checks

To define groups for built-in health checks, add the following grouping hint to your config.yml:

    default_group: default
            default: # checks you may want to execute by default
                php_extensions: [apc, xdebug]
            cron: # checks you may want to execute only on cron servers
                php_extensions: [redis]

This creates two groups, default and cron, each having different checks.

Define groups for tagged Services

To define groups for tagged services, add a group attribute to the respective tags:

        class: Acme\HelloBundle\Check\PhpExtensionsCheck
            - [ xhprof, apc, memcache ]
            - { name: liip_monitor.check, alias: php_extensions, group: cron }
            - { name: liip_monitor.check, alias: php_extensions, group: app_server }

autoconfigure will place checks into the default group. You must add autoconfigure: false to the service definition to change the group:

        autoconfigure: false
            - { name: liip_monitor.check, group: app_server }

Specify group for CLI commands

Both CLI commands have a --group=... option. If it is not given, the default group is used.

bin/console monitor:list --group=app_server
bin/console monitor:health --group=app_server

Both commands, monitor:list and monitor:health, have an option --all to list or run the checks of all registered groups. Additionally, the monitor:list has an option --groups to list all registered groups.

Full Default Config

    enable_controller:    false
    view_template:        null
    failure_status_code:  502
        enabled:              false
        recipient:            ~ # Required
        sender:               ~ # Required
        subject:              ~ # Required
        send_on_warning:      true
    default_group:        default

        # Grouping checks

            # Prototype

                # Validate that a named extension or a collection of extensions is available
                php_extensions:       [] # Example: apc, xdebug

                # Pairs of a PHP setting and an expected value
                php_flags:            # Example: session.use_only_cookies: false

                    # Prototype
                    setting:              ~

                # Pairs of a version and a comparison operator
                php_version:          # Example: 5.4.15: >=

                    # Prototype
                    version:              ~

                # Process name/pid or an array of process names/pids
                process_running:      ~ # Example: [apache, foo]

                # Validate that a given path (or a collection of paths) is a dir and is readable
                readable_directory:   [] # Example: ["%kernel.cache_dir%"]

                # Validate that a given path (or a collection of paths) is a dir and is writable
                writable_directory:   [] # Example: ["%kernel.cache_dir%"]

                # Validate that a class or a collection of classes is available
                class_exists:         [] # Example: ["Lua", "My\Fancy\Class"]

                # Benchmark CPU performance and return failure if it is below the given ratio
                cpu_performance:      ~ # Example: 1.0 # This is the power of an EC2 micro instance

                # Checks to see if the disk usage is below warning/critical percent thresholds
                    warning:              70
                    critical:             90
                    path:                 '%kernel.cache_dir%'

                # Checks Symfony2 requirements file
                    file:                 '%kernel.root_dir%/SymfonyRequirements.php'

                # Checks to see if the OpCache memory usage is below warning/critical thresholds
                    warning:              70
                    critical:             90

                # Checks to see if the APC memory usage is below warning/critical thresholds
                    warning:              70
                    critical:             90

                # Checks to see if the APC fragmentation is below warning/critical thresholds
                    warning:              70
                    critical:             90

                # Connection name or an array of connection names
                doctrine_dbal:        null # Example: [default, crm]
                # Checks to see if migrations from specified configuration file are applied
                    # Examples:
                        configuration_file:  %kernel.root_dir%/Resources/config/migrations.yml
                        connection:          default
                        connection:          default
                    migrations_with_doctrine_bundle_v2: default
                    # Prototype
                        # Absolute path to doctrine migrations configuration
                        configuration_file:   ~
                        # Connection name from doctrine DBAL configuration
                        connection:           ~ # Required
                # Connection name or an array of connection names
                doctrine_mongodb:        null # Example: [default, crm]

                # Check if MemCache extension is loaded and given server is reachable

                    # Prototype
                        host:                 localhost
                        port:                 11211

                # Validate that a Redis service is running

                    # Prototype
                        host:                 localhost
                        port:                 6379
                        password:             null
                        # or
                        dsn: redis://localhost:6379

                # Attempt connection to given HTTP host and (optionally) check status code and page content

                    # Prototype
                        host:                 localhost
                        port:                 80
                        path:                 /
                        status_code:          200
                        content:              null

                # Attempt connection using Guzzle to given HTTP host and (optionally) check status code and page content

                    # Prototype
                        url:                  localhost
                        headers:              []
                        options:              []
                        status_code:          200
                        content:              null
                        method:               GET
                        body:                 null

                # Validate that a RabbitMQ service is running

                    # Prototype
                        host:                 localhost
                        port:                 5672
                        user:                 guest
                        password:             guest
                        vhost:                /
                        # or
                        dsn: amqp://guest:guest@localhost:5672/%2F

                # Checks the version of this app against the latest stable release
                symfony_version:      ~

                # Checks if error pages have been customized for given error codes
                    # The status codes that should be customized
                    error_codes:          [] # Required

                    # The directory where your custom error page twig templates are located. Keep as "%kernel.project_dir%" to use default location.
                    path:                 '%kernel.project_dir%'

                # Checks installed composer dependencies against the SensioLabs Security Advisory database
                    lock_file:            '%kernel.root_dir%/../composer.lock'

                # Validate that a stream wrapper or collection of stream wrappers exists
                stream_wrapper_exists:  [] # Example: ['zlib', 'bzip2', 'zip']

                # Find and validate INI files
                file_ini:             [] # Example: ['path/to/my.ini']

                # Find and validate JSON files
                file_json:            [] # Example: ['path/to/my.json']

                # Find and validate XML files
                file_xml:             [] # Example: ['path/to/my.xml']

                # Find and validate YAML files
                file_yaml:            [] # Example: ['path/to/my.yml']
                # PDO connections to check for connection

                    # Prototype
                        dsn:                  null
                        username:             null
                        password:             null
                        timeout:              1

                # Checks that fail/warn when given expression is false (expressions are evaluated with symfony/expression-language)

                    # Example:
                        label:               OPcache
                        warning_expression:  ini('opcache.revalidate_freq') > 0
                        critical_expression: ini('opcache.enable')
                        warning_message:     OPcache not optimized for production
                        critical_message:    OPcache not enabled

                    # Prototype
                        label:                ~ # Required
                        warning_expression:   null # Example: ini('apc.stat') == 0
                        critical_expression:  null # Example: ini('short_open_tag') == 1
                        warning_message:      null
                        critical_message:     null

                # Validate that a messenger transport does not contain more than warning/critical messages
                # Transport must implement MessageCountAwareInterface
                    name: # name of transport
                        critical_threshold:   10   # required
                        warning_threshold:    null # optional: warning level
                        service:              null # defaults to 


For documentation on the REST API see: Don't forget to add the bundle routes in your routing.xml file.

Nagios integration

You can find a simple Nagios check written in Perl and Python in the Resources/scripts directory.

Perl Version

This is dependent on perl modules available on CPAN Getopt::Std, WWW::Mechanize, and JSON

Copy the script into your scripts directory in Nagios and create a command like this:

define command{
        command_name    check_symfony_health
        command_line    $USER1$/ -H $HOSTNAME$

Running the command with the Hostname flag (-H) will check "http://$HOSTNAME$/monitor/health/run". You can also use the Address flag (-A) to check a specified URL:

command_line    $USER1$/ -A

The plugin can be used with Authentication, Using the Username (-u) and Password (-p) flags:

command_line    $USER1$/check_symfony2.p1 -H $HOSTNAME$ -u username -p password

You can also specify the Warning (-w) and Critical (-c) levels for the check using the standard flags

command_line    $USER1$/ -H $HOSTNAME$ -w 1 -c 2

Any flags can be combined except -A and -H. THe -u and -p flags should always be used together.

Python Version

The Python version depends on the nagiosplugin library < 1.0.0.

Copy the script into your scripts directory in Nagios and create a command like this:

define command{
        command_name    check_symfony_health
        command_line    $USER1$/ -w 0  -c 0 -u https://$HOSTNAME$

To use the plugin with HTTP basic authentication, change the command to:

command_line    $USER1$/ -w 0  -c 0 -u https://$HOSTNAME$ -a username:password

Connecting Check to Host in Nagios

Add a service:

define service{
 hostgroup_name         Symfony2
 service_description    Symfony2 health check
 check_command          check_symfony_health
 use                    generic-service

And create a host attached to the Symfony2 hostgroup:

define host{
    use              web-host
    hostgroups       Symfony2

And place your host within the Symfony2 hostgroup.