Applies to SUSE OpenStack Cloud 9

25 DNS Service Installation Overview #

The SUSE OpenStack Cloud DNS Service supports several different backends for domain name service. The choice of backend must be included in the deployment model before the SUSE OpenStack Cloud install is completed.

Warning

By default any user in the project is allowed to manage a DNS domain. This can be changed by updating the Policy.json file for designate.

The backends that are available within the DNS Service are separated into two categories, self-contained and external.

Table 25.1: DNS Backends #

Category	Backend	Description	Recommended For
Self-contained	BIND 9.x	All components necessary will be installed and configured as part of the SUSE OpenStack Cloud install.	POCs and customers who wish to keep cloud and traditional DNS separated.
External	InfoBlox	The authoritative DNS server itself is external to SUSE OpenStack Cloud. Management and configuration is out of scope for the Cloud Lifecycle Manager but remains the responsibility of the customer.	Customers who wish to integrate with their existing DNS infrastructure.

25.1 Installing the DNS Service with BIND #

SUSE OpenStack Cloud DNS Service defaults to the BIND back-end if another back-end is not configured for domain name service. BIND will be deployed to one or more control planes clusters. The following configuration example shows how the BIND service is installed.

25.1.1 Configuring the Back-end #

Ensure the DNS Service components and the BIND component have been placed on a cluster. BIND can be placed on a cluster separate from the other DNS service components.

control-planes:
          - name: control-plane-1
          region-name: region1

          clusters:
          - name: cluster1
          service-components:
          - lifecycle-manager
          - mariadb
          - ip-cluster
          - apache2
          - ...
          - designate-api
          - designate-central
          - designate-pool-manager
          - designate-zone-manager
          - designate-mdns
          - designate-client
          - bind

Updating the Input Model

When the back-end is configured, add bind-ext to the file network_groups.yml.

Edit ~/openstack/my_cloud/definition/data/network_groups.yml to add bind-ext to component-endpoints.
```
name: EXTERNAL-API
hostname-suffix: extapi
component-endpoints:
- bind-ext
```
Save the file.

25.2 Install the DNS Service with PowerDNS #

25.2.1 Installing DNS Service with PowerDNS #

SUSE OpenStack Cloud DNS Service and PowerDNS can be installed together instead of the default BIND backend. PowerDNS will be deployed to one or more control planes clusters. The following configuration example shows how the PowerDNS service is installed.

25.2.2 Configure the Backend #

To configure the backend for PowerDNS, follow these steps.

Ensure the DNS Service components and the PowerDNS component have been placed on a cluster. PowerDNS may be placed on a separate cluster to the other DNS Service components. Ensure the default bind component has been removed.

control-planes:
          - name: control-plane-1
          region-name: region1

          clusters:
          - name: cluster1
          service-components:
          - lifecycle-manager
          - mariadb
          - ip-cluster
          - apache2
          - ...
          - designate-api
          - designate-central
          - designate-pool-manager
          - designate-zone-manager
          - designate-mdns
          - designate-client
          - powerdns

Edit the ~/openstack/my_cloud/definitions/data/network_groups.yml file to include the powerdns-ext.

- name: EXTERNAL-API
hostname-suffix: extapi
component-endpoints:
 - powerdns-ext
load-balancers:
 - provider: ip-cluster

Edit the ~/openstack/my_cloud/definitions/data/firewall_rules.yml to allow UDP/TCP access.

	    - name: DNSudp
      # network-groups lists the network group names that the rules apply to
      network-groups:
      - EXTERNAL-API
      rules:
      - type: allow
        # range of remote addresses in CIDR format that this rule applies to
        remote-ip-prefix:  0.0.0.0/0
        port-range-min: 53
        port-range-max: 53
        protocol: udp

    - name: DNStcp
      # network-groups lists the network group names that the rules apply to
      network-groups:
      - EXTERNAL-API
      rules:
      - type: allow
        # range of remote addresses in CIDR format that this rule applies to
        remote-ip-prefix:  0.0.0.0/0
        port-range-min: 53
        port-range-max: 53
        protocol: tcp

Please see Section 10.3.2, “designate Initial Configuration” for post-installation DNS Service configuration.

25.3 Configure DNS Domain and NS Records #

To configure the default DNS domain and Name Server records for the default pool, follow these steps.

Ensure that designate_config.yml file is present in the ~/openstack/my_cloud/definition/data/designate folder. If the file or folder is not present, create the folder and copy designate_config.yml file from one of the example input models (for example, ~/openstack/examples/entry-scale-kvm/data/designate/designate_config.yml).

Modify the dns_domain and/or ns_records entries in the designate_config.yml file.

data:
dns_domain: example.org.
ns_records:
    hostname: ns1.example.org.
    priority: 1
    hostname: ns2.example.org.
    priority: 2

Edit your input model's control_plane.yml file to include DESIGNATE-CONFIG-CP1 in configuration-data section.

control-planes:
   - name: control-plane-1
     region-name: region1
     lifecycle-manager-target
     configuration-data:
        - DESIGNATE-CONFIG-CP1
        - NEUTRON-CONFIG-CP1

Continue your cloud deployment by reviewing and committing your changes.

$ git add ~/openstack/my_cloud/definition/data/designate/designate_config.yml
$ git commit -m "Adding DNS Domain and NS Records"

Note

In an entry-scale model (Section 9.3.1, “Entry-Scale Cloud”), you will have 3 ns_records since the DNS service runs on all three control planes.

In a mid-scale model (Section 9.3.3, “Single-Region Mid-Size Model”) or dedicated metering, monitoring and logging model (Section 9.3.2, “Entry Scale Cloud with Metering and Monitoring Services”), the above example would be correct since there are only two controller nodes.

25.4 Migrate Zone/Pool to Worker/Producer after Upgrade #

After upgrade, the following steps may be used to migrate the zone/pool scheme to the worker/producer scheme:

Edit the file ~/openstack/my_cloud/definitions/data/control_plane.yml to replace designate-pool-manager to designate-worker and designate-zone-manager to designate-producer

control-planes:
          - name: control-plane-1
          region-name: region1

          clusters:
          - name: cluster1
          service-components:
          - lifecycle-manager
          - mariadb
          - ip-cluster
          - apache2
          - ...
          - designate-api
          - designate-central
          - designate-worker
          - designate-producer
          - designate-mdns
          - designate-client
          - powerdns

Commit your changes to git:

ardana > cd ~/openstack/ardana/ansible
ardana > git commit -m "COMMIT_MESSAGE"

Run the configuration processor:

ardana > cd ~/openstack/ardana/ansible
ardana > ansible-playbook -i hosts/localhost config-processor-run.yml

Create a deployment directory:

ardana > cd ~/openstack/ardana/ansible
ardana > ansible-playbook -i hosts/localhost ready-deployment.yml

Run the designate-migrate.yml playbook to migrate the designate services:

ardana > cd ~/scratch/ansible/next/ardana/ansible
ardana > ansible-playbook -i hosts/verb_hosts designate-migrate.yml