07. Monitoring Review & Practice

Pro Tips on Using the ELK Stack

The ELK stack requires more memory than everything else in this course.

• The standard config recommends 2GB. This is why we migrate to a larger instance.
• It may also be necessary to stop anything else that you have set up in this course (i.e. Jenkins, Prometheus, and Grafana).
• If the Ansible installation script for Elasticsearch takes too long (more than 5 minutes for a single item) you may have run out of memory.
• Do not be alarmed, just restart the EC2 instance either through the command line, or if necessary in the AWS web console.

Further Reading on Monitoring

To learn how to set up alerts and notifications with ELK stack, please refer to the following:

• https://www.elastic.co/what-is/elasticsearch-alerting

Practice your Skills with Monitoring! - Practice Exercise 1

• Migrate your AWS instance to a T2.medium
• Set up Prometheus and configure a node exporter
  https://github.com/cloudalchemy/ansible-prometheus
  curl -LO https://s3-us-west-2.amazonaws.com/grafana-releases/release/grafana_5.1.4_am
• Install Grafana and configure it to connect to your Prometheus backend
• Take a screenshot of your Grafana graph, and compare it to mine below

Practice Exercise 2

• Set up ELK stack and all of its components (note: upgrade the version of Elasticsearch in Ansible, or do an apt install logstash to match those for Logstash and Kibana)
  https://github.com/elastic/ansible-elasticsearch
• Now compare your screen of Kibana showing logs from your system to mine below. They should be similar.