health-monitor-rails

This is a health monitoring Rails mountable plug-in, which checks various services (db, cache, sidekiq, redis, etc.).

Mounting this gem will add a '/check' route to your application, which can be used for health monitoring the application and its various services. The method will return an appropriate HTTP status as well as an HTML/JSON/XML response representing the state of each provider.

You can filter which checks to run by passing a parameter called providers.

Examples

HTML Status Page

JSON Response

>> curl -s http://localhost:3000/check.json | json_pp

{
   "timestamp" : "2017-03-10 17:07:52 +0200",
   "status" : "ok",
   "results" : [
      {
         "name" : "Database",
         "message" : "",
         "status" : "OK"
      },
      {
         "status" : "OK",
         "message" : "",
         "name" : "Cache"
      },
      {
         "status" : "OK",
         "message" : "",
         "name" : "Redis"
      },
      {
         "status" : "OK",
         "message" : "",
         "name" : "Sidekiq"
      }
   ]
}

Filtered JSON Response

>> curl -s http://localhost:3000/check.json?providers[]=database&providers[]=redis | json_pp

{
   "timestamp" : "2017-03-10 17:07:52 +0200",
   "status" : "ok",
   "results" : [
      {
         "name" : "Database",
         "message" : "",
         "status" : "OK"
      },
      {
         "status" : "OK",
         "message" : "",
         "name" : "Redis"
      },
   ]
}

XML Response

>> curl -s http://localhost:3000/check.xml

<?xml version="1.0" encoding="UTF-8"?>
<hash>
  <results type="array">
    <result>
      <name>Database</name>
      <message></message>
      <status>OK</status>
    </result>
    <result>
      <name>Cache</name>
      <message></message>
      <status>OK</status>
    </result>
    <result>
      <name>Redis</name>
      <message></message>
      <status>OK</status>
    </result>
    <result>
      <name>Sidekiq</name>
      <message></message>
      <status>OK</status>
    </result>
  </results>
  <status type="symbol">ok</status>
  <timestamp>2017-03-10 17:08:50 +0200</timestamp>
</hash>

Filtered XML Response

>> curl -s http://localhost:3000/check.xml?providers[]=database&providers[]=redis

<?xml version="1.0" encoding="UTF-8"?>
<hash>
  <results type="array">
    <result>
      <name>Database</name>
      <message></message>
      <status>OK</status>
    </result>
    <result>
      <name>Redis</name>
      <message></message>
      <status>OK</status>
    </result>
  </results>
  <status type="symbol">ok</status>
  <timestamp>2017-03-10 17:08:50 +0200</timestamp>
</hash>

Setup

If you are using bundler add health-monitor-rails to your Gemfile:

gem 'health-monitor-rails'

Then run:

bundle install

Otherwise, install the gem:

gem install health-monitor-rails

Usage

You can mount this inside your app routes by adding this to config/routes.rb:

mount HealthMonitor::Engine, at: '/'

Supported Service Providers

The following services are currently supported:

• DB
• Cache
• Redis
• Sidekiq
• Resque
• Delayed Job
• Solr
• FileAbsence

Configuration

Adding Providers

By default, only the database check is enabled. You can add more service providers by explicitly enabling them via an initializer:

HealthMonitor.configure do |config|
  config.cache
  config.redis
  config.sidekiq
  config.delayed_job
end

We believe that having the database check enabled by default is very important, but if you still want to disable it (e.g., if you use a database that isn't covered by the check) - you can do that by calling the no_database method:

HealthMonitor.configure do |config|
  config.no_database
end

Provider Configuration

All providers accept a general set of baseline configuration:

HealthMonitor.configure do |config|
  config.[provider].configure do |provider_config|
    provider_config.name = 'Redis'
    provider_config.critical = true
  end
end

• name: Custom name for the provider (Defaults to class name. Ex: 'Redis', 'Sidekiq')
• critical: Whether or not the provider is a critical dependency (Defaults to: true). If set to false, the monitor will report its status but ignore it when determining overall application health status

The critical option allows you to monitor for additional non-critical dependencies that are not fully required for your application to be operational, like a cache database for instance

Some of the providers can also accept additional configuration:

# Sidekiq
HealthMonitor.configure do |config|
  config.sidekiq.configure do |sidekiq_config|
    sidekiq_config.latency = 3.hours
    sidekiq_config.queue_size = 50
  end
end

# Sidekiq with specific queues
HealthMonitor.configure do |config|
  config.sidekiq.configure do |sidekiq_config|
    sidekiq_config.add_queue_configuration('critical', latency: 10.seconds, queue_size: 20)
  end
end

# Redis with existing connection
HealthMonitor.configure do |config|
  config.redis.configure do |redis_config|
    redis_config.connection = Redis.current # Use your custom redis connection
    redis_config.max_used_memory = 200 # Megabytes
  end
end

Additionally, you can configure an explicit URL:

# Redis with a URL configuration
HealthMonitor.configure do |config|
  config.redis.configure do |redis_config|
    redis_config.url = 'redis://user:[email protected]:90210/'
    redis_config.max_used_memory = 200
  end
end

Or via a connection pool:

# Redis using Connection Pools
HealthMonitor.configure do |config|
  config.redis.configure do |redis_config|
    redis_config.connection = ConnectionPool.new(size: 5) { Redis.new } # Use your custom connection pool
  end
end

For providers that can be configured with its endpoits/urls you can also add multiple declarations to ensure you are reporting across all dependencies:

HealthMonitor.configure do |config|
  config.redis.configure do |c|
    c.name = 'Redis: Cache'
    c.url = ENV.fetch('REDISCLOUD_URL', 'redis://localhost:6379/0')
  end
  config.redis.configure do |c|
    c.name = 'Redis: Action Cable'
    c.url = ENV.fetch('REDISCLOUD_URL', 'redis://localhost:6379/0')
  end
  config.redis.configure do |c|
    c.name = 'Redis: Sidekiq'
    c.url = ENV.fetch('REDISCLOUD_URL', 'redis://localhost:6379/1')
  end
end

The currently supported settings are:

Sidekiq

• latency: the latency (in seconds) of a queue (now - when the oldest job was enqueued) which is considered unhealthy (the default is 30 seconds, but larger processing queue should have a larger latency value).
• queue_size: the size (maximum) of a queue which is considered unhealthy (the default is 100).
• default_queue: the default queue to check.
• add_queue_configuration: add specific configuration per queue.

Redis

• url: the URL used to connect to your Redis instance. Note, that this is an optional configuration and will use the default connection if not specified. You can also use url to explicitly configure authentication (e.g., 'redis://user:[email protected]:90210/').
• connection: Use custom Redis connection (e.g., Redis.current).
• max_used_memory: Set maximum expected memory usage of Redis in megabytes. Prevent memory leaks and keys over storing.

Please note that url or connection can't be used at the same time.

Delayed Job

• queue_size: the size (maximum) of a queue which is considered unhealthy (the default is 100).

Solr

• url: the URL used to connect to your Solr instance - must be a string. You can also use url to explicitly configure authentication (e.g., 'https://user:[email protected]:8983/')
• collection: An optional parameter used to connect to your specific Solr collection - must be a string. By setting this parameter the code will check the status of this individual collection in your Solr instance instead of just the status of the overall Solr instance

FileAbsence

This check allows you to create a file on your server when you would like to force the check to fail. For example if you are utilizing the health.json as you health check page for your load balancer and would like to force a machine offline.

• filename: the file relative to the rails root that must remain absent for the health check to remain passing. For example: public/remove-from-nginx

Adding a Custom Provider

It's also possible to add custom health check providers suited for your needs (of course, it's highly appreciated and encouraged if you'd contribute useful providers to the project).

To add a custom provider, you'd need to:

• Implement the HealthMonitor::Providers::Base class and its check! method (a check is considered as failed if it raises an exception):

class CustomProvider < HealthMonitor::Providers::Base
  def check!
    raise 'Oh oh!'
  end
end

• Add its class to the configuration:

HealthMonitor.configure do |config|
  config.add_custom_provider(CustomProvider)
end

Adding a Custom Error Callback

If you need to perform any additional error handling (for example, for additional error reporting), you can configure a custom error callback:

HealthMonitor.configure do |config|
  config.error_callback = proc do |e|
    logger.error "Health check failed with: #{e.message}"

    Raven.capture_exception(e)
  end
end

Adding Authentication Credentials

By default, the /check endpoint is not authenticated and is available to any user. You can authenticate using HTTP Basic Auth by providing authentication credentials:

HealthMonitor.configure do |config|
  config.basic_auth_credentials = {
    username: 'SECRET_NAME',
    password: 'Shhhhh!!!'
  }
end

Adding Environment Variables

By default, environment variables are nil, so if you'd want to include additional parameters in the results JSON, all you need is to provide a Hash with your custom environment variables:

HealthMonitor.configure do |config|
  config.environment_variables = {
    build_number: 'BUILD_NUMBER',
    git_sha: 'GIT_SHA'
  }
end

Customizing the path

By default, the endpoint where the status page is served is /check, but this can be customized:

HealthMonitor.configure do |config|
  config.path = :status
end

This will make the page to be served in the /status endpoint for instance (from where the engine was mounted).

Monitoring Script

A Nagios/Shinken/Icinga/Icinga2 plugin is available in extra directory.

It takes one argument: -u or --uri

nicolas@desktop:$ ./check_rails.rb
missing argument: uri

Usage: check_rails.rb -u uri
    -u, --uri URI                    The URI to check (https://nagios:[email protected]/check.json)

Common options:
    -v, --version                    Displays Version
    -h, --help                       Displays Help

Also, it generates an output with the right status code for your monitoring system:

nicolas@desktop:$ ./check_rails.rb -u http://admin:admin@localhost:5000/check.json
Rails application : OK

Database : OK
Cache : OK
Redis : OK
Sidekiq : OK

nicolas@desktop:$ echo $?
0

nicolas@desktop:$ ./check_rails.rb -u http://admin:admin@localhost:5000/check.json
Rails application : ERROR

Database : OK
Cache : OK
Redis : ERROR (Error connecting to Redis on 127.0.0.1:6379 (Errno::ECONNREFUSED))
Sidekiq : ERROR (Error connecting to Redis on 127.0.0.1:6379 (Errno::ECONNREFUSED))

nicolas@desktop:$ echo $?
2

Development

In order to work on development on the gem itself

Installing the gems

Use the appraisal gem to install the bundles for different rails versions:

appraisal clean
appraisal generate
appraisal install

Running the tests

Use appraisal to run the tests using rake

appraisal rake

License

The MIT License (MIT)

Copyright (c) 2017

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.