We strongly recommend storing logs in one centralized place when working in a clustered environment. MongooseIM uses Lager - the logging framework. Its backend can be easily replaced; the syslog backend is included by default in MongooseIM.

Using syslog

To activate the syslog backend you have to edit rel/files/app.config and uncomment the line:

%% use below line to add syslog backend for Lager
%        {lager_syslog_backend, [ "mongooseim", local0, info]},

Remember to provide a parameter list to make your lager syslog backend running:

  • The first parameter is a string to tag all the syslog messages with. The default is mongooseim.
  • The second one is the facility to log to (see the syslog documentation).
  • The last parameter is the lager level at which the backend accepts messages. In our case it's info.

Depending on the system platform you use, remember also to add the appropriate line in the syslog config file:

local0.info                     /var/log/mongooseim.log

Now all the logs of level info will be passed to the /var/log/mongooseim.log file.

Example log (e.g tail -f /var/log/mongooseim.log):

Apr  1 12:36:49 User.local mongooseim[6068]: [info] <0.7.0> Application mnesia started on node mongooseim@localhost

Further / multiserver integration

For more advanced processing and analysis of logs, including gathering logs from multiple machines, you can use one of the many available systems (e.g. logstash/elasticsearch/kibana, graylog, splunk), which collect data from the syslog and are beyond the scope of this documentation.



WombatOAM is an operations and maintenance framework for Erlang based systems. Its Web Dashboard displays this data in an aggregated manner and provides interfaces to feed the data to other OAM tools such as Graphite, Nagios or Zabbix.

For more information see: WombatOAM.


To monitor MongooseIM during load testing, we recommend the following open source applications:

  • Graphite is used for data presentation.
  • collectd is a daemon running on the monitored nodes capturing data related to CPU and Memory usage, IO etc.


It provides a REST interface for Mongoose's metrics, so it can be easily integrated with other services.

You can read more about it here: REST interface to metrics.

Built-in Exometer reporters

MongooseIM uses the Exometer libary for collecting the metrics. Exometer has many build-in reporters that can send metrics to external services like:

  • graphite
  • amqp
  • statsd
  • snmp
  • opentsdb

It is possible to enable them in MoongooseIM via the app.config file. The file sits next to the ejabberd.cfg file in the rel/files and _REL_DIR_/etc directories. For more details, please visit the Exometer's project page: Exometer.

Note that we are using the 1.2.1 version.

Below you can find a sample configuration. It shows setting up a reporter connecting to graphite running on localhost.

You can see an additional option not listed in the Exometer docs - mongooseim_report_interval. That option sets the metrics resolution: how often Exometer gathers and sends metrics through reporters. By default that is 60 seconds.

{exometer, [
    {mongooseim_report_interval, 60000}, %% 60 seconds
    {report, [
        {reporters, [
                     {exometer_report_graphite, [
                                                 {prefix, "mongooseim"},
                                                 {connect_timeout, 5000},
                                                 {host, ""},
                                                 {port, 2003},
                                                 {api_key, ""}

Run graphite in Docker - quick start

Start a docker machine:

docker-machine start

Make sure it is running:

$ docker-machine status

Run the following command:

$ docker run -d --name graphite --restart=always hopsoft/graphite-statsd

Get the "local" IP address of the container:

$ docker inspect graphite | grep IPAdd | grep -v Secon | cut -d '"' -f 4 | head -n 1

Get IP address of the machine:

$ docker-machine ip

Route subnet:

$ sudo route add -net

And now should show a graphite page.

Check if the data collection works - run:

$ while true; do echo -n "example:$((RANDOM % 100))|c" | nc -w 1 -u 8125; done

Wait a while, then open:

Then, if you configure your MongooseIM to send Exometer reports to that IP address and run it for a while, you should be able to see some interesting charts.