shpaz / openshift-infra-roles

Openshift Infrastructure Ansible Roles

This repository is used for deploying various Openshift infrastructure workloads using Ansible Automation.

Docker V2 Registry

The registry role will create a Docker V2 registry using ansible local connection. This approach is very useful for mirroring images before a disconnected containerized installation.

Consul

The Consul role will create a Consul cluster which will react as Load-balancer, service discovery, DNS resolver, etc. This Consul cluster will point to the RGW instances to load the traffic intelligently.

This rule pulls a Consul container image from a given location, extracts the Consul binary and created the wanted cluster. Important! don't forge to set the consul_interface so that Ansible will know which port it should bind the traffic to.

To run the automation:

• You should change the dictionary located in group_vars/consul-servers.yml to your RGW instances, Ansible will load them as consul config
• Run the command bash ansible-playbook playbooks/consul.yml -i hosts --extra-vars "registry_username= registry_password="

This will create a three-node Consul cluster, and will serve the UI in port 80. To access your S3 service, use the following url - http://s3.service.dc1.domain1 for example.

S3bench

The S3bench role will create an entire infrastructure for you to start benchmarking your S3 service in a multiclient manner. This role creates a pod that interacts with your S3 service, created a workload and documents all results in an ELK stack deployed by this role as well. In addition, There is an automation for throwing in some pre-defined dashboards that will help you analyze the results you get when performing the workload.

This role divides into three parts:

• The first part, deploys the ELK stack, and can be initiated by runinng ansible-playbook playbooks/s3bench.yml -i hosts --tags start_infra. This will deploy an ELK stack, that will be available by host networking and can be accessed via port 5601 (Kibana port).
• The second part, deploys the s3bench service, which is being defined by a workload located in group_vars/s3bench.yml. Once you edit this vars file with your configuration (such as endpoint_url, access_key, secret_key, bucket_name, etc) you could pick the hosts you want to run these on by editing the inventory file. After running ansible-playbook playbooks/s3bench.yml -i hosts --tags start_s3bench, the playbook will add the needed containers to the pod and start testing your S3 service.
• The third part, created the dashboards, running ansible-playbook playbooks/s3bench.yml -i hosts --tags create_dashboards will send an API request to Kibana wit hthe needed ndjson file.