Elasticsearch, Logstash and Kibana 5.6.16

This is a small container at only 300Mb compressed, running a full functional ELK 5 stack.

Important

Make sure your docker hosts has the folowing sysctl setting, this is required for ELK

insert in /etc/sysctl.conf

vm.max_map_count = 262144

or run

sysctl -w vm.max_map_count=262144

Features

• filebeat support
• cisco syslog support
• yum.log support via filebeat
• nginx accesslogs support
• updated upstream grok patterns
• running on Alpine Linux with s6, small, clean and efficient
• Maxmind geo data enabled
• Each process runs as own user, in docker ;)
• multi input index is created based on type

Instructions

Start the container

docker run -d -p 5601:5601 -p 9200:9200 -p 5044:5044 \
  -v /var/lib/elasticsearch:/var/lib/elasticsearch \
  --name elk \
  cdrocker/elk5:latest

Check progress with

docker logs -f elk

You can now open kibana http://elasticsearchhost:5601

There will probably be no index patterns, you'll have to import them manually. For beats you can use the new import_dashboards script which automate this process. (Install filebeat for this functionality.)

/usr/share/filebeat/scripts/import_dashboards -es http://<elasticsearch>:9200
/usr/share/metricbeat/scripts/import_dashboards -es http://<elasticsearch>:9200
/usr/share/packetbeat/scripts/import_dashboards -es http://<elasticsearch>:9200

Todo

• Add java environment options
• autoupdate GEO data
• curator install
• auto cleanup of old indices
• elasticsearch plugins