Logs Explorer for Docker aka LED

Purpose

The goal is to provide an easy process for exploring and visualizing container logs.

LED wants to keep everything as simple as possible. But it is in very early stage

The following picture shows you a quick look of a running LED instance.

LED is designed for microservice architecture builds with docker.

If you are already familiar with elastic stack, LED understanding will be very easy for you.

Logging Driver

LED assumes that you are using docker logging driver to send container logs to the Fluentd collector as structured log data.

How to use it ?

1- Launch container

LED requires the following ports to be published:

• 8080 : used by the hosted web server to serves led
• 24224: default port used by fluentd for TCP forwarding

 docker run -d -p 8080:8080 -p 24224:24224 bwnyasse/led:0.7.0

Navigate to localhost:8080 to see a basic running instance of LED.

2- Add container logs

The following command will connect the docker hello-world container to LED. ( See the official documentation )

docker run -d \
    --name=hello_world_container \
    --log-driver=fluentd \
    --log-opt tag="default.docker.{{.Name}}" \
    hello-world

Navigate to localhost:8080 to see your container logs in LED

That's it !!

Important : Setting the Timezone in a Docker image

Time in your running container may be out of sync with your host and can cause you troubles when exploring your logs.
To avoid this behavoir , one of the good pratices is to read-only mount your host /etc/timezone and /etc/localtime in your container.

Launching led as follow :

docker run -d \
      -v /etc/localtime:/etc/localtime:ro \
      -v /etc/timezone:/etc/timezone:ro \
      -p 8080:8080 \
      -p 24224:24224 \
      bwnyasse/led:0.7.0

Connecting a mysql database as follow :

docker run -d \
    -v /etc/localtime:/etc/localtime:ro \
    -v /etc/timezone:/etc/timezone:ro \
    --name=mysql_db \
    -e MYSQL_ROOT_PASSWORD=my-secret-pw \
    --log-driver=fluentd \
    --log-opt tag="default.docker.{{.Name}}" \
    mysql

Customize your running instance of LED

- Using Environment variables

Example:

Assuming LED , that we have the following requirements:

• Wolverine as Application Name
• LED will run under a reverse proxy and will be available to MY_HOST:MY_PORT
• We want to trigger elastic search curation every 5 minutes

    docker run -d \
          -v /etc/localtime:/etc/localtime:ro \
          -v /etc/timezone:/etc/timezone:ro \
          -p 8080:8080 \
          -p 24224:24224 \
          -e APP_NAME=Wolverine \
          -e APP_CONTEXT_URL=MY_HOST:MY_PORT \
          -e ES_CURATOR_SCHEDULE=*/5 * * * *
          bwnyasse/led:0.7.0

- Level color

Use the settings page, to modify the level color. The following picture give you a quick look of current available level Configuration

- Backing up data

• Logs in LED ElasticSearch will be store in /opt/elasticsearch/data
• LED configuration will be store in /opt/led/conf

For more info check Docker docs section on [Managing data in containers(https://docs.docker.com/userguide/dockervolumes/)

Manage Tag option

Sometimes , it is more better to filter logs by level. At this stage of project, LED display log level only with the default log format of the following services :

• Jboss Wildfly & Wildfly Swarm
• MongoDB

The following table show you fluent tag requires by LED to match service and if the log level is displayed or not

Connecting wildlfy as follow :

docker run -d \
    -v /etc/localtime:/etc/localtime:ro \
    -v /etc/timezone:/etc/timezone:ro \
    --name=wildfly_server_sync \
    --log-driver=fluentd \
    --log-opt tag="wildfly.docker.{{.Name}}" \
    jboss/wildfly:10.0.0.Final

Connecting MongoDB as follow :

docker run -d \
    -v /etc/localtime:/etc/localtime:ro \
    -v /etc/timezone:/etc/timezone:ro \
    --name=mongo_db \
    --log-driver=fluentd \
    --log-opt tag="mongo.docker.{{.Name}}" \
    mongo:3.2.8

More to come ... ( see issues ) :