Using a custom domain
Every application running on the Cloud Platform is able to use a hostname for its HTTP endpoints, using a pre-defined domain:
This works automatically, using a wildcard TLS certificate.
Most applications will need their own, application-specific
These hostnames (or usually, entire domains) are managed individually and a
number of actions are required to set them up.
Domains are managed inside DNS zones. You can read more about the structure of the Domain Name System in this page. It is recommended that applications use their own DNS zones, to provide better isolation, and make management simpler.
Defining the DNS zone
To simplify management, we recommend that you only define a single zone, as part of the resources for your production environment, if possible. You can still use subdomains of that zone for different environments.
Once the zone is created, you will need to setup the necessary name server (NS) records in the parent DNS zone, before you can use it. For more information on how this delegation method works, you can read about authoritative name servers in this page.
If your zone is for a subdomain of
the Cloud Platform team can help you set it up; please create a support ticket.
For any other domains (including any subdomain of
https://myservice.gov.uk), you will need to contact the parent zone’s
administrators (usually GDS) to set this up. If in doubt, don’t hesitate to get
in touch with us in
Please note, once you setup the NS records, you’ll be delegating control of the
zone to the Cloud Platform. Hostnames used by your services (using
will be automatically managed by the cluster.
Obtaining a certificate
- Create the Certificate resource, filling in any placeholders with your
secretNameattribute defines the name of a kubernetes secret in your namespace where the certificate and key material will be stored.
Please add the certificate by defining it in your namespace folder in the environments repository. Doing so means it will automatically be restored in the event of any problems with our certificate/DNS systems.
--- apiVersion: cert-manager.io/v1alpha3 kind: Certificate metadata: name: <my-cert> namespace: <my-namespace> spec: secretName: <my-cert-secret> issuerRef: name: letsencrypt-production kind: ClusterIssuer dnsNames: - <hostname>
- Make sure the certificate has been issued correctly, by checking its
$ kubectl describe certificate <my-cert>
The certificate status of type
"Ready" should be
Status: Conditions: Last Transition Time: 2019-06-05T10:16:43Z Message: Certificate is up to date and has not expired Reason: Ready Status: True Type: Ready Not After: 2019-09-03T09:16:42Z Events: Type Reason Age From Message ---- ------ ---- ---- ------- Normal Generated 3m cert-manager Generated new private key Normal OrderCreated 3m cert-manager Created Order resource "<my-cert>-3189350212" Normal OrderComplete 1m cert-manager Order "<my-cert>-3189350212" completed successfully Normal CertIssued 1m cert-manager Certificate issued successfully
It generally takes just a few minutes for the certificate to be prepared and
the events displayed should indicate if there is a problem, or it simply needs
more time. If you cannot obtain a certificate, please get in touch with us in
- You will need to update your
Ingressspec to include the new hostname.
apiVersion: networking.k8s.io/v1beta1 kind: Ingress metadata: name: <ingress-name> namespace: <namespace-name> annotations: kubernetes.io/ingress.class: nginx external-dns.alpha.kubernetes.io/set-identifier: <ingress-name>-<namespace-name>-<colour> external-dns.alpha.kubernetes.io/aws-weight: "100" spec: tls: - hosts: - my-app.apps.live-1.cloud-platform.service.justice.gov.uk + - hosts: + - <hostname> + secretName: <my-cert-secret> rules: - host: my-app.apps.live-1.cloud-platform.service.justice.gov.uk http: paths: - path: / backend: serviceName: <my-svc> servicePort: 80 + - host: <hostname> + http: + paths: + - path: / + backend: + serviceName: <my-svc> + servicePort: 80
(Note - Please change the
namespace-name values in the above example. The
colour should be
blue for ingress in live-1 and
green for ingress in EKS
Once you’ve made the changes to your
Ingress, the cluster (and more
external-dns) will update the necessary records defined in it.
This usually takes less than a minute until you are able to access your
endpoint. However, depending on the DNS name servers your workstation uses, you
might need to wait longer or try to “flush” your local DNS cache in order to
speed up the process. You should search online for the proper method to do so,
based on your operating system and/or browser.