Configuring Logstash for Dockeredit

Logstash differentiates between two types of configuration: Settings and Pipeline Configuration.

Pipeline Configurationedit

It is essential to place your pipeline configuration where it can be found by Logstash. By default, the container will look in /usr/share/logstash/pipeline/ for pipeline configuration files.

In this example we use a bind-mounted volume to provide the configuration via the docker run command:

docker run --rm -it -v ~/pipeline/:/usr/share/logstash/pipeline/ docker.elastic.co/logstash/logstash:6.3.2

Every file in the host directory ~/pipeline/ will then be parsed by Logstash as pipeline configuration.

If you don’t provide configuration to Logstash, it will run with a minimal config that listens for messages from the Beats input plugin and echoes any that are received to stdout. In this case, the startup logs will be similar to the following:

Sending Logstash logs to /usr/share/logstash/logs which is now configured via log4j2.properties.
[2016-10-26T05:11:34,992][INFO ][logstash.inputs.beats    ] Beats inputs: Starting input listener {:address=>"0.0.0.0:5044"}
[2016-10-26T05:11:35,068][INFO ][logstash.pipeline        ] Starting pipeline {"id"=>"main", "pipeline.workers"=>4, "pipeline.batch.size"=>125, "pipeline.batch.delay"=>5, "pipeline.max_inflight"=>500}
[2016-10-26T05:11:35,078][INFO ][org.logstash.beats.Server] Starting server on port: 5044
[2016-10-26T05:11:35,078][INFO ][logstash.pipeline        ] Pipeline main started
[2016-10-26T05:11:35,105][INFO ][logstash.agent           ] Successfully started Logstash API endpoint {:port=>9600}

This is the default configuration for the image, defined in /usr/share/logstash/pipeline/logstash.conf. If this is the behaviour that you are observing, ensure that your pipeline configuration is being picked up correctly, and that you are replacing either logstash.conf or the entire pipeline directory.

Settingsedit

The image provides several methods for configuring settings. The conventional approach is to provide a custom logstash.yml file, but it’s also possible to use environment variables to define settings.

Bind-mounted settings filesedit

Settings files can also be provided through bind-mounts. Logstash expects to find them at /usr/share/logstash/config/.

It’s possible to provide an entire directory containing all needed files:

docker run --rm -it -v ~/settings/:/usr/share/logstash/config/ docker.elastic.co/logstash/logstash:6.3.2

Alternatively, a single file can be mounted:

docker run --rm -it -v ~/settings/logstash.yml:/usr/share/logstash/config/logstash.yml docker.elastic.co/logstash/logstash:6.3.2

Bind-mounted configuration files will retain the same permissions and ownership within the container that they have on the host system. Be sure to set permissions such that the files will be readable and, ideally, not writeable by the container’s logstash user (UID 1000).

Custom Imagesedit

Bind-mounted configuration is not the only option, naturally. If you prefer the Immutable Infrastructure approach, you can prepare a custom image containing your configuration by using a Dockerfile like this one:

FROM docker.elastic.co/logstash/logstash:6.3.2
RUN rm -f /usr/share/logstash/pipeline/logstash.conf
ADD pipeline/ /usr/share/logstash/pipeline/
ADD config/ /usr/share/logstash/config/

Be sure to replace or delete logstash.conf in your custom image, so that you don’t retain the example config from the base image.

Environment variable configurationedit

Under Docker, Logstash settings can be configured via environment variables. When the container starts, a helper process checks the environment for variables that can be mapped to Logstash settings. Settings that are found in the environment are merged into logstash.yml as the container starts up.

For compatibility with container orchestration systems, these environment variables are written in all capitals, with underscores as word separators

Some example translations are shown here:

Table 1. Example Docker Environment Variables

Environment Variable

Logstash Setting

PIPELINE_WORKERS

pipeline.workers

LOG_LEVEL

log.level

XPACK_MONITORING_ENABLED

xpack.monitoring.enabled

In general, any setting listed in the settings documentation can be configured with this technique.

Defining settings with environment variables causes logstash.yml to be modified in place. This behaviour is likely undesirable if logstash.yml was bind-mounted from the host system. Thus, it is not reccomended to combine the bind-mount technique with the environment variable technique. It is best to choose a single method for defining Logstash settings.

Docker defaultsedit

The following settings have different default values when using the Docker images:

http.host

0.0.0.0

xpack.monitoring.elasticsearch.url

http://elasticsearch:9200

The setting xpack.monitoring.elasticsearch.url is not defined in the -oss image.

These settings are defined in the default logstash.yml. They can be overridden with a custom logstash.yml or via environment variables.

If replacing logstash.yml with a custom version, be sure to copy the above defaults to the custom file if you want to retain them. If not, they will be "masked" by the new file.

Logging Configurationedit

Under Docker, Logstash logs go to standard output by default. To change this behaviour, use any of the techniques above to replace the file at /usr/share/logstash/config/log4j2.properties.