The purpose of this project is to provide a simple, yet flexible deployment of OpenShift on OpenStack using a three step process. This guide assumes you are familiar with OpenStack.
If you want to provide additional features, please feel free to contribute via pull requests or any other means. We are happy to track and discuss ideas, topics and requests via 'Issues'.
For each release of OpenShift a release branch will be created. Starting with OpenShift 3.9 we will follow the OpenShift release version so it is easy to tell what release branch goes with OpenShift version. The installer support OpenShift Enterprise and OKD starting with 3.11. Note: CentOS OKD rpms are released after enterprise so you may have to wait a bit for OKD.
- release-3.9 OpenShift 3.9
- release-3.10 OpenShift 3.10
- release-3.11 OpenShift 3.11
In addition I would like to metion I borrowed a lot of ideas from two other projects.
- Working OpenStack deployment. Tested is OpenStack 12 & 13 (Pike & Queens) using RDO.
- RHEL 7 image. Tested is RHEL 7.4, 7.5, 7.6.
- An openstack ssh key for accessing instances
- A provider (public) network with at least two or three available floating ips.
- A service (internal) network
- A router configured with the public network as gateway and internal network as interface.
- Flavors configured for OpenShift. These are only recommendations.
- ocp.master (2 vCPU, 4GB RAM, 30 GB Root Disk)
- ocp.infra (4 vCPU, 16GB RAM, 30 GB Root Disk)
- ocp.node (2 vCPU, 4GB RAM, 30 GB Root Disk)
- ocp.bastion (1 vCPU, 4GB RAM, 30 GB Root Disk)
- Properly configured cinder and nova storage.
- Make sure you aren't using default loop back and have disabled disk zeroing in cinder/nova for LVM.
More information on setting up proper OpenStack environment can be found here.
Single Master - Non HA
Single Master deployment is 1 Master, 1 Infra node and X number of App nodes. This configuration is a non-HA setup, ideal for test environments.
Multiple Master - HA
Multiple Master deployment is 3 Master, 2 Infra node and X number of App nodes. This configuration is an HA setup. By default etcd and registry are not using persistent storage. This would need to be configured post-install manually at this time if those should be persisted.
Setup New OpenStack Project
Create Project
# openstack project create openshift
Add admin user as admin to project
# openstack role add --project openshift --user admin admin
Increase project quota for security groups
# openstack quota set --secgroups 100 openshift
Increase quota for volumes
# openstack quota set --volumes 100 openshift
Create internal network
# openstack network create openshift --project openshift
Create internal subnet
# openstack subnet create --network openshift --allocation-pool \
start=192.168.4.100,end=192.168.4.200 --dns-nameserver 213.133.98.98 \
--subnet-range 192.168.4.0/24 openshift_subnet
Add internal network to router as interface
# openstack router add subnet router1 openshift_subnet
This deployment will configure authentication through OpenStack keystone. This means you need to create users in OpenStack so they are available to OpenShift. All users that successfully authenticate from OpenShift to Keystone will be allowed to login. They will only have permissions to create new projects and not see anyone elses projects. If you would like something else you can configure the inventory file manually before deploying OpenShift cluster.
Create OpenStack User for OpenShift
# openstack user create --project admin --password <password> <username>
[OpenStack Controller]
Install Git & Ansible
# yum install -y git ansible
Clone Git Repository
# git clone https://github.com/ktenzer/openshift-on-openstack-123.git
Change dir to repository
# cd openshift-on-openstack-123
Checkout release branch 3.11
# git checkout release-3.11
Configure Parameters
# cp sample-vars.yml vars.yml
# vi vars.yml
---
### OpenStack Setting ###
openstack_user: admin
openstack_passwd: <password>
openstack_auth_url: <openstack auth url>
openstack_project: openshift
domain_name: ocp3.lab
external_network: public
service_network: openshift
service_subnet_id: <openshift_subent id>
image: rhel76
ssh_user: cloud-user
ssh_key_path: /root/admin.pem
ssh_key_name: admin
stack_name: openshift
openstack_release: queens
openstack_version: "13"
contact: [email protected]
heat_template_path: /root/openshift-on-openstack-123/heat/openshift_single_lbaas.yaml
### OpenShift Settings ###
openshift_deployment: openshift-enterprise
openshift_version: "3.11"
docker_version: "1.13.1"
openshift_ha: false
registry_replicas: 1
openshift_user: admin
openshift_passwd: <password>
### Red Hat Subscription ###
subscription_use_username: True
rhn_username_or_org_id: <user or org id>
rhn_password_or_activation_key: <password or activation key>
rhn_pool: <pool>
### OpenStack Instance Count ###
master_count: 1
infra_count: 1
node_count: 3
### OpenStack Instance Group Policies ###
### Set to 'anti-affinity' if running on multiple compute node ###
master_server_group_policies: "['soft-anti-affinity']"
infra_server_group_policies: "['soft-anti-affinity']"
node_server_group_policies: "['soft-anti-affinity']"
### OpenStack Instance Flavors ###
bastion_flavor: ocp.bastion
master_flavor: ocp.master
infra_flavor: ocp.infra
node_flavor: ocp.node
Note: If you want to run a single load balancer (to save floating ips) for masters and infra, instead of default two use following heat template heat_template_path: /root/openshift-on-openstack-123/heat/openshift_single_lbaas.yaml.
# ./01_deploy-openstack-infra.yml --private-key=<path to openstack key for accessing instances>
Get ip address of the bastion host.
# openstack stack output show openshift --all | grep -A1 '"name": "bastion"'
| "name": "bastion",
| "address": "1.2.3.4"
SSH to the bastion host using cloud-user and key.
ssh -i /root/admin.pem [email protected]
[Bastion Host]
Change dir to repository
# cd openshift-on-openstack-123
[cloud-user@bastion ~]$ ./02_prepare-openshift.yml
PLAY RECAP *****************************************************************************************
bastion : ok=15 changed=7 unreachable=0 failed=0
infra0 : ok=18 changed=13 unreachable=0 failed=0
infra1 : ok=18 changed=13 unreachable=0 failed=0
localhost : ok=7 changed=6 unreachable=0 failed=0
master0 : ok=18 changed=13 unreachable=0 failed=0
master1 : ok=18 changed=13 unreachable=0 failed=0
master2 : ok=18 changed=13 unreachable=0 failed=0
node0 : ok=18 changed=13 unreachable=0 failed=0
node1 : ok=18 changed=13 unreachable=0 failed=0
[Bastion Host]
[cloud-user@bastion ~]$ ansible-playbook -i /home/cloud-user/openshift-inventory -vv /usr/share/ansible/openshift-ansible/playbooks/prerequisites.yml
PLAY RECAP *****************************************************************************************
infra0.ocp3.lab : ok=61 changed=15 unreachable=0 failed=0
localhost : ok=11 changed=0 unreachable=0 failed=0
master0.ocp3.lab : ok=73 changed=15 unreachable=0 failed=0
node0.ocp3.lab : ok=61 changed=15 unreachable=0 failed=0
INSTALLER STATUS ***********************************************************************************
Initialization : Complete (0:04:16)
[cloud-user@bastion ~]$ ansible-playbook -i /home/cloud-user/openshift-inventory -vv /usr/share/ansible/openshift-ansible/playbooks/deploy_cluster.yml
PLAY RECAP *****************************************************************************************
infra0 : ok=118 changed=20 unreachable=0 failed=0
localhost : ok=11 changed=0 unreachable=0 failed=0
master0 : ok=715 changed=237 unreachable=0 failed=0
node0 : ok=118 changed=21 unreachable=0 failed=0
INSTALLER STATUS ***********************************************************************************
Initialization : Complete (0:00:32)
Health Check : Complete (0:00:01)
Node Bootstrap Preparation : Complete (0:09:24)
etcd Install : Complete (0:01:06)
Master Install : Complete (0:06:04)
Master Additional Install : Complete (0:01:51)
Node Join : Complete (0:00:37)
Hosted Install : Complete (0:00:49)
Cluster Monitoring Operator : Complete (0:01:15)
Web Console Install : Complete (0:00:31)
Console Install : Complete (0:00:27)
metrics-server Install : Complete (0:00:00)
Service Catalog Install : Complete (0:01:57)
Login in to UI.
https://openshift.144.76.134.226.xip.io:8443
OKD formally called OpenShift Origin (community version) is also supported starting with release-3.11 branch. To use OKD make sure you have a centos 7.5 image and set 'openshift_deployment=origin' in the vars file.
Once you have run the 01_deploy-openstack-infra.yml and 03_prepare-openshift.yml playbooks as documented above run the following to install openshift OKD from bastion.
Prerequisites playbook
[centosr@bastion ~] ansible-playbook -i /home/centos/openshift-inventory openshift-ansible/playbooks/prerequisites.yml
Deploy cluster playbook
[centosr@bastion ~] ansible-playbook -i /home/centos/openshift-inventory openshift-ansible/playbooks/deploy_cluster.yml
Configure admin user
[cloud-user@bastion ~]$ ssh master0
Authenticate as system:admin user.
[cloud-user@master0 ~]$ oc login -u system:admin -n default
Make user OpenShift Cluster Administrator
[cloud-user@master0 ~]$ oc adm policy add-cluster-role-to-user cluster-admin admin
Note: Metrics is integrated with OpenShift UI and will be depricated in 4.0 but for 3.11 if you want metrics in UI it is still needed.
Set metrics to true in inventory
[cloud-user@bastion ~]$ vi openshift_inventory
...
openshift_hosted_metrics_deploy=true
...
Run playbook for metrics in OpenShift
[cloud-user@bastion ~]$ ansible-playbook -i /home/cloud-user/openshift-inventory -vv /usr/share/ansible/openshift-ansible/playbooks/openshift-metrics/config.yml
PLAY RECAP *****************************************************************************************
infra0.ocp3.lab : ok=0 changed=0 unreachable=0 failed=0
localhost : ok=11 changed=0 unreachable=0 failed=0
master0.ocp3.lab : ok=217 changed=47 unreachable=0 failed=0
node0.ocp3.lab : ok=0 changed=0 unreachable=0 failed=0
INSTALLER STATUS ***********************************************************************************
Initialization : Complete (0:01:34)
Metrics Install : Complete (0:04:37)
Set logging to true in inventory
[cloud-user@bastion ~]$ vi openshift_inventory
...
openshift_hosted_logging_deploy=true
...
Run Playbook for logging in OpenShift
[cloud-user@bastion ~]$ ansible-playbook -i /home/cloud-user/openshift-inventory -vv /usr/share/ansible/openshift-ansible/playbooks/openshift-logging/config.yml
[cloud-user@bastion ~]$ vi openshift_inventory
...
openshift_additional_registry_credentials=[{'host':'registry.connect.redhat.com','user':'<your_user_name>','password':'<your_password>','test_image':'mongodb/enterprise-operator:0.3.2'}]
...
Reconfigure registry auth
[cloud-user@bastion ~]$ ansible-playbook -i /home/cloud-user/openshift-inventory -vv /usr/share/ansible/openshift-ansible/playbooks/updates/registry_auth.yml
PLAY RECAP ********************************************************************************************************************************************************************************************************
infra0 : ok=27 changed=1 unreachable=0 failed=0
infra1 : ok=27 changed=1 unreachable=0 failed=0
localhost : ok=13 changed=0 unreachable=0 failed=0
master0 : ok=48 changed=2 unreachable=0 failed=0
master1 : ok=48 changed=2 unreachable=0 failed=0
master2 : ok=65 changed=2 unreachable=0 failed=0
node0 : ok=27 changed=1 unreachable=0 failed=0
node1 : ok=27 changed=1 unreachable=0 failed=0
node2 : ok=27 changed=1 unreachable=0 failed=0
INSTALLER STATUS **************************************************************************************************************************************************************************************************
Initialization : Complete (0:03:08)
Deploy Operator Framework
[cloud-user@bastion ~]$ ansible-playbook -i /home/cloud-user/openshift-inventory -vv /usr/share/ansible/openshift-ansible/playbooks/olm/config.yml
PLAY RECAP ********************************************************************************************************************************************************************************************************
infra0 : ok=0 changed=0 unreachable=0 failed=0
infra1 : ok=0 changed=0 unreachable=0 failed=0
localhost : ok=11 changed=0 unreachable=0 failed=0
master0 : ok=27 changed=0 unreachable=0 failed=0
master1 : ok=27 changed=0 unreachable=0 failed=0
master2 : ok=68 changed=19 unreachable=0 failed=0
node0 : ok=0 changed=0 unreachable=0 failed=0
node1 : ok=0 changed=0 unreachable=0 failed=0
node2 : ok=0 changed=0 unreachable=0 failed=0
INSTALLER STATUS **************************************************************************************************************************************************************************************************
Initialization : Complete (0:01:30)
OLM Install : Complete (0:00:47)
Disconnected installation requires two Swift containers, which are used to store Docker images and RHEL repositories.
01_deploy-openstack-infra.yml playbook will do the following:
- Install httpd on Bastion server
- Install and configure Rclone, mount
openshift_rhn_repocontainer via systemd mount script to/var/www/html/repo - Generate CA and SSL certificate for Docker private registry.
- Setup Docker and configure Docker private registry to use Swift container
openshift_rhn_registryas backend storage. - Replicate latest RHEL packages from required repositories (6GB), and sync them to
openshift_rhn_repocontainer. - Download all required Openshift Docker images (6GB), re-tag them and push to private Docker registry running on Bastion.
note: replication of RHEL repositories and Openshift Docker images will only happen once, if data exists in Swift containers this steps will be skipped.
Set variables in vars.yml
bastion_repo: True
Create two required Swift containers
swift_rhn_repo_container_name: openshift_rhn_repo
swift_rhn_registry_container_name: openshift_rhn_registry
Run below playbook after running 01_deploy-openstack-infra.yml
# ./02_bastion-repo.yml
Continue with step 2.
Currently using the OpenStack cloud provider requires using Cinder v2 API. Most current OpenStack deployments will default to v3.
Error creating cinder volume: BS API version autodetection failed.
If you provision OpenShift volume and it is pending check /var/log/messages on master. If you see this error you need to add following in /etc/origin/cloudprovider/openstack.conf on masters and all nodes then restart node service on node and controller service on master.
...
[BlockStorage]
bs-version=v2
...
The post-openshift.yml playbook takes care of setting v2 for cinder automatically.
The registry sometimes fails to complete install due to host resolution of xip.io. Not sure if this is issue in OpenShift 3.7 or environment. Simply re-running hosted playbook resolved the issue and resulted in successful installation.
Don't use ansible 2.7 with OpenShift or OKD 3.11 there is an issue where etcd playbooks won't run which leads to control plane not starting. Use 2.6, this is tested and working. This deployment will force 2.6 so 2.7 doesnt end up on systems by accident.