Skip to content

Latest commit

 

History

History
70 lines (61 loc) · 3.9 KB

README.md

File metadata and controls

70 lines (61 loc) · 3.9 KB
Raw

FAKE Monitoring Service

This repo contains a HELM Chart capable of spinning up a cluster consisting of the following:

This allows us to do the following:

  • Collect and parse logs across all our pods and the internal kubernetes logs (Fluent Bit)
  • Collect metrics across various application (APM Server)
  • Collect metrics from the K8S cluster
  • Track users through the the various APIs (APM Server)
  • Allow various dahsboards to be build (Kibana)
  • Allow reports to be generated (Kibana)
  • Allow the features above to interoperate and if needed to be correlated together (Elasticsearch)

Project structure

The project consists of a HELM Chart which described the interaction and configuration of various services. To make our service more portable in Rancher We add a couple of files described here.

file purpose
app-read.md Decorates the Rancher App page
Chart.yaml Contains various meta data describing our service
docker-compose.yml Contains an example of the service usable by a docker-compose commando
README.md Contains the description of the service visible in Rancher
requirements.lock Contains the locked down versions specified in the file below
requirements.yaml Contains Helm Chart this service is dependendant on
values.yaml Contains configuration values

Local Development

Due to the complex nature of Kuberenetes not everything can be properly tested locally, to at least attempt at running it locally you can do the following:

  • Enable Kubernetes in your docker daemon
  • Install K8S Helm for helm file support
    • brew install kubernetes-helm
  • Initialize Tiller on the local K8S
    • helm init
  • Add elastic repo for the latest helm files containing elasticsearch and kibana but not apm-server :(
    • helm repo add elastic https://helm.elastic.co
  • Update dependencies by using in the monitoring servcice folder
    • helm dependency update
  • Install the service on your local K8S cluster using the name fake
    • helm install --debug ./monitoring-service --name fake
  • Expose the port for Kibana on your host machine
    • kubectl port-forward <kibana-docker-id> 5601
  • Delete the service to free up resources
    • helm del --purge fake

Rancher Interaction

To use our fancy new service as an app in your fancy new Rancher setup you have to do the following:

  • Add the github repo as a Rancher Catalog (Do note that Bitbucket repo's don't seem to work due to some internal rancher issue)
  • Launch the app
  • Change the Kibana Ingress to the preferred value

Future work

This is not production ready due to the following:

  • ElasticSearch data is not persisted due to disabled settings
  • ElasticSearch is not using a username/password
  • Kibana is accessible by everyone
  • Resources might be on the low side for a production version
  • RDAC and roles are not used but should be
  • All data is shipped to ElasticSearch, some should be redacted due to privacy concerns
  • Logstash format might be incomplete due to some settings
  • Questions.yaml file is missing and could be used to specify Ingress settings before spinning everything up in Rancher
  • templates/_helper.tpl file is missing and could be used to guide a user to the running Kibana instance
  • The Other README.md file is lacking in up to date information