MongoDB logger for Rails
Ruby HTML CSS CoffeeScript Other
Latest commit 102ee48 Dec 15, 2016 @le0pard committed on GitHub Merge pull request #71 from pedrogarrett/master
Use a thread variable instead of an instance var, to keep threa…
Failed to load latest commit information.
app/assets move to exe Apr 6, 2015
examples refactoring of js code Dec 2, 2012
exe move to exe Apr 6, 2015
gemfiles new rails 4.2 and ruby 2.2.0 Dec 29, 2014
lib Use a thread variable instead of an instance var, to allow keep threa… Dec 14, 2016
spec Support uncapped collections Jan 7, 2016
tasks typo Oct 7, 2013
.gitignore support for non rails applications Jan 29, 2014
.ruby-version ruby Dec 27, 2015
.travis.yml ruby Dec 27, 2015
Appraisals new rails 4.2 and ruby 2.2.0 Dec 29, 2014 bump 0.6.5 Dec 29, 2014
Gemfile change gemspec Nov 27, 2012
LICENSE Add license Nov 14, 2011 Mention uncapped collections in README Jan 7, 2016
Rakefile cleanup Mar 1, 2014 fix configs Mar 1, 2014


Build Status Code Climate

MongodbLogger is a alternative logger for Rails or Rack based app, which log all requests of you application into MongoDB database. It:

  • simple to integrate into existing Rails application;
  • allow to store all logs from web cluster into one scalable storage - MongoDB;
  • flexible schema of MongoDB allow to store and search any information from logs;
  • web panel allow filter logs, build graphs using MapReduce by information from logs;

Rails support

Please note the latest version is compatible with rails 3.1.x or newer.

For rails 3.0.x latest version 0.2.8.

Doesn't support the Rails version below 3.


  1. Add the following to your Gemfile then refresh your dependencies by executing "bundle install" (or just simple "bundle"):

    gem "mongodb_logger"
  2. Add adapter in Gemfile. Supported mongo and moped (mongoid). For example:

    gem "mongo"
    gem "bson_ext"


      gem "moped"
  3. Add the following line to your ApplicationController:

    include MongodbLogger::Base
  4. For use with Heroku you need to prevent the rails_log_stdout plugin from being added by Heroku for rails 3:

    mkdir vendor/plugins/rails_log_stdout
    touch vendor/plugins/rails_log_stdout/.gitkeep

    For Rails 4 just remove from Gemfile "rails_12factor" gem.

  5. Add MongodbLogger settings to database.yml for each environment in which you want to use the MongodbLogger. The MongodbLogger will also look for a separate mongodb_logger.yml or mongoid.yml (if you are using mongoid) before looking in database.yml. In the mongodb_logger.yml and mongoid.yml case, the settings should be defined without the 'mongodb_logger' subkey.


      adapter: postgresql
      database: my_app_development
      username: postgres
        database: my_app               # required (the only required setting)
        capped: true                   # default: true  - warning: uncapped collections introduce the vulnerability that the size of the collection grows too high, exceeding all avaialble disk space
        capsize: <%= 10.megabytes %>   # default: 250MB - ignored if capped is set to false
        host: localhost                # default: localhost
        port: 27017                    # default: 27017
        username: null                 # default: null, username for MongoDB Auth
        password: null                 # default: null, password for MongoDB Auth
        replica_set: true              # default: false - Adds retries for ConnectionFailure during voting for replica set master
        write_options:                 # default: {w: 0, wtimeout: 200} - write options for inserts (w - wait for insert to propagate to "w" numbers of nodes)
          w: 0
          wtimeout: 200
        application_name: my_app       # default: Rails.application
        disable_file_logging: false    # default: false - disable logging into filesystem (only in MongoDB)
        collection: some_name          # default: Rails.env + "_log" - name of MongoDB collection


      database: my_app
      capsize: <%= 10.megabytes %>
      host: localhost
      port: 27017
      replica_set: true

    Also you can use "url" parameter for setup connection to mongodb:

      url: mongodb://username:password@localhost:27017/my_app
      capsize: <%= 10.megabytes %>
  6. For using with MongoDB Replica Set (more info you can read by this link In config set list of [host, port] in key "hosts":

      database: my_app
      capsize: <%= 10.megabytes %>
      host: localhost
      port: 27017
        - -
          - 27018
        - -
          - 27019
  7. For assets pipeline you can generate all js/css file into folder by rake task:

    rake mongodb_logger:assets:compile[public/assets]

Assets pipeline

For capistrano possible compile assets by receipt. Add this to config/deploy.rb:

require 'mongodb_logger/capistrano'
set :mongodb_logger_assets_dir, "public/assets" # where to put mongodb assets
after 'deploy:update_code', 'mongodb_logger:precompile'

Also you can serve assets from rails app. You need just mount it separately:

mount, :at => "/mongodb", :as => :mongodb
mount MongodbLogger::Assets.instance, :at => "/mongodb/assets", :as => :mongodb_assets # assets


After success instalation of gem, a new MongoDB document (record) will be created for each request on your application, by default will record the following information: Runtime, IP Address, Request Time, Controller, Method, Action, Params, Application Name and All messages sent to the logger. The structure of the MongoDB document looks like this:

  'action'           : action_name,
  'application_name' : application_name (rails root),
  'controller'       : controller_name,
  'ip'               : ip_address,
  'messages'         : {
                         'info'  : [ ],
                         'debug' : [ ],
                         'error' : [ ],
                         'warn'  : [ ],
                         'fatal' : [ ]
  'params'           : { },
  'path'             : path,
  'request_time'     : date_of_request,
  'runtime'          : elapsed_execution_time_in_milliseconds,
  'url'              : full_url,
  'method'           : request method (GET, POST, OPTIONS),
  'session'          : information from session,
  'is_exception'     : true only for exceptions (in other cases this field miss)

Beyond that, if you want to add extra information to the base of the document (let's say something like user_id on every request that it's available), you can just call the Rails.logger.add_metadata method on your logger like so (for example from a before_filter):

# make sure we're using the MongodbLogger in this environment
Rails.logger.add_metadata(user_id: if Rails.logger.respond_to?(:add_metadata)

Callback on exceptions

For send email or do something on exception you can add callback:

MongodbLogger::Base.configure do |config|
  config.on_log_exception do |mongo_record|
    # do something with this data, for example - send email (better - by background job)

In this callback send record without "_id", because logger not wait for insert response from MongoDB.

Disable mongodb_logger

To disable MongodbLogger you can use option disable. But this should be set before rails load (for example in Rails app the top of "config/application.rb"):

MongodbLogger::Base.configure do |config|
  config.disable = true


 MongodbLogger::Base.disable = true

Migrate to another size of capped collection

If you need change capper collection size, you should change the "capsize" key in mongodb_config and run this task for migration:

rake mongodb_logger:migrate

Rack Middleware

If you want use MongodbLogger in Rack app which is mounted to your Rails app, you can try to use rack middleware:

use MongodbLogger::RackMiddleware


If you want use MongodbLogger with some of Rails::Engine, you can do this (example for Spree):

Spree::BaseController.send :include, MongodbLogger::Base
Spree::Admin::BaseController.send :include, MongodbLogger::Base # for admin

The Front End

To setup web interface in you Rails application, first of all create autoload file in you Rails application

File: you_rails_app/config/initializers/mongodb_logger.rb (example)

require 'mongodb_logger/server' # required
# this secure you web interface by basic auth, but you can skip this, if you no need this
MongodbLogger::Server.use Rack::Auth::Basic do |username, password|
    [username, password] == ['admin', 'password']

and just mount MongodbLogger::Server in rails routes:

File: you_rails_app/config/routes.rb

mount, :at => "/mongodb"

Now you can see web interface by url "http://localhost:3000/mongodb"

If you've installed MongodbLogger as a gem and want running the front end without Rails application, you can do it by this command:

mongodb_logger_web config.yml

where config.yml is config, similar to config of Rails apps, but without Rails.env. Example:

database: app_logs_dev
host: localhost
port: 27017
collection: development_log # set for see development logs

parameter "collection" should be set, if your set custom for your Rails application or start this front end not for production enviroment (by default taken "production_log" collection, in Rails application gem generate "#{Rails.env}_log" collection, if it is not defined in config).

It's a thin layer around rackup so it's configurable as well:

mongodb_logger_web config.yml -p 8282

Passenger, Unicorn, Thin, etc.

Using Passenger, Unicorn, Thin, etc? MongodbLogger ships with a you can use. See guide:

Don't forget setup MONGODBLOGGERCONFIG env variable, which provide information about MongodbLogger config. Example starting with unicorn:

MONGODBLOGGERCONFIG=examples/server_config.yml unicorn

Demo Application with MongodbLogger


Demo Sources:

Querying via the Rails console

And now, for a couple quick examples on getting ahold of this log data... First, here's how to get a handle on the MongoDB from within a Rails console:

>> db = Rails.logger.mongo_adapter.connection
=> #<Mongo::DB:0x007fdc7c65adc8 @name="monkey_logs_dev" ... >
>> collection = Rails.logger.mongo_adapter.collection
=> #<Mongo::Collection:0x007fdc7a4d12b0 @name="development_log" .. >

Once you've got the collection, you can find all requests for a specific user (with id):

>> cursor = collection.find(:user_id => '12355')
=> #<Mongo::Cursor:0x1031a3e30 ... >
>> cursor.count
=> 5

Find all requests that took more that one second to complete:

>> collection.find({:runtime => {'$gt' => 1000}}).count
=> 3

Find all order#show requests with a particular order id (id=order_id):

>> collection.find({"controller" => "order", "action"=> "show", "" => order_id})

Find all requests with an exception that contains "RoutingError" in the message or stack trace:

>> collection.find({"messages.error" => /RoutingError/})

Find all requests with errors:

>> collection.find({"is_exception" => true})

Find all requests with a request_date greater than '11/18/2010 22:59:52 GMT'

>> collection.find({:request_time => {'$gt' => Time.utc(2010, 11, 18, 22, 59, 52)}})

Using Uncapped Collections as Storage

MongoDB's capped collections are a safe choice for storing logs where expiration happens automatically after exceeding the provided capsize (first in first out expiration).

Capped collections comes with a few limitations, one of them being that you cannot manually delete log entries. Switching to a capped: false configuration will store all log entries in an uncapped collection and remove the constraints of uncapped collections.

Warning: If you choose to deploy mongodb_logger with an uncapped collection configuration, you should implement an alternative way of cleaning up log records (cron job or similar). Uncapped collections can grow indifinely in size and take up more disk space than you anticipated.

Copyright (c) 2009-2014 Phil Burrows, CustomInk (based on and Leopard released under the MIT license