Prerequisites

• An ingress provides network access for backend workloads. Ensure that a workload is available in a cluster. If no workload is available, deploy a workload by referring to Creating a Deployment, Creating a StatefulSet, or Creating a DaemonSet.
• A Service has been configured for the workload. Table 1 lists the supported Service types. For details about how to configure the Service, see Intra-Cluster Access (ClusterIP) or NodePort.

  ENI LoadBalancer refers to the passthrough networking backed by using a CCE Turbo cluster and a dedicated ELB load balancer.

• Dedicated load balancers must be the application type (HTTP/HTTPS) supporting private networks (with a private IP).

Precautions

• It is recommended that other resources not use the load balancer automatically created by an ingress. Otherwise, the load balancer will be occupied when the ingress is deleted, resulting in residual resources.
• After an ingress is created, upgrade and maintain the configuration of the selected load balancers on the CCE console. Do not modify the configuration on the ELB console. Otherwise, the ingress service may be abnormal.
• The URL registered in an ingress forwarding policy must be the same as the URL exposed by the backend Service. Otherwise, a 404 error will be returned.
• In a cluster using the IPVS proxy mode, if the ingress and Service use the same ELB load balancer, the ingress cannot be accessed from the nodes and containers in the cluster because kube-proxy mounts the LoadBalancer Service address to the ipvs-0 bridge. This bridge intercepts the traffic of the load balancer connected to the ingress. You are advised to use different ELB load balancers for the ingress and Service.

Adding an ELB Ingress

This section uses an Nginx workload as an example to describe how to add an ELB ingress.

1. Log in to the CCE console and access the cluster details page.
2. Choose Networking in the navigation pane, click the Ingresses tab, and click Create Service in the upper right corner.
3. Set ingress parameters.
   • Name: Specify a name of an ingress, for example, ingress-demo.
   • Namespace: Select the namespace to which the ingress is to be added.
   • Interconnect with Nginx: This parameter is displayed only after the nginx-ingress add-on is installed. It is disabled by default. If this function is enabled, Nginx Ingress Controller will be used. For details, see Using Nginx Ingresses on the Console.
   • Load Balancer

     Select the load balancer to interconnect. Only load balancers in the same VPC as the cluster are supported. If no load balancer is available, click Create Load Balancer to create one on the ELB console.

     Dedicated load balancers must support HTTP and the network type must support private networks.

   • Listener Configuration: Ingress configures a listener for the load balancer, which listens to requests from the load balancer and distributes traffic. After the configuration is complete, a listener is created on the load balancer. The default listener name is k8s__<Protocol type>_<Port number>, for example, k8s_HTTP_80.
     • Front-End Protocol: HTTP and HTTPS are available.
     • External Port: Port number that is open to the ELB service address. The port number can be specified randomly.
     • Server Certificate: When an HTTPS listener is created for a load balancer, you need to bind a certificate to the load balancer to support encrypted authentication for HTTPS data transmission. For details on how to create a secret, see Creating a Secret.
       

       If there is already an HTTPS ingress for the chosen port on the load balancer, the certificate of the new HTTPS ingress must be the same as the certificate of the existing ingress. This means that a listener has only one certificate. If two certificates, each with a different ingress, are added to the same listener of the same load balancer, only the certificate added earliest takes effect on the load balancer.

     • SNI: Click to enable the Server Name Indication (SNI) function. SNI is an extended protocol of TLS. It allows multiple TLS-based access domain names to be provided for external systems using the same IP address and port number. Different domain names can use different security certificates. After SNI is enabled, the client is allowed to submit the requested domain name when initiating a TLS handshake request. After receiving the TLS request, the load balancer searches for the certificate based on the domain name in the request. If the certificate corresponding to the domain name is found, the load balancer returns the certificate for authorization. Otherwise, the default certificate (server certificate) is returned for authorization.
       

       • The SNI option is available only when HTTPS is selected.

       • This function is supported only for clusters of v1.15.11 and later.
       • Specify the domain name for the SNI certificate. Only one domain name can be specified for each certificate. Wildcard-domain certificates are supported.
     • Security Policy: combinations of different TLS versions and supported cipher suites available to HTTPS listeners.

       For details about security policies, see Security Policy.

       

       • Security Policy is available only when HTTPS is selected.
       • This function is supported only for clusters of v1.17.9 and later.
   • Forwarding Policies: When the access address of a request matches the forwarding policy (a forwarding policy consists of a domain name and URL, for example, 10.117.117.117:80/helloworld), the request is forwarded to the corresponding target Service for processing. Click Add Forwarding Policies to add multiple forwarding policies.
     • Domain Name: actual domain name. Ensure that the domain name has been registered and archived. Once a domain name rule is configured, you must use the domain name for access.
     • Rule Matching
       • Prefix match: If the URL is set to /healthz, the URL that meets the prefix can be accessed. For example, /healthz/v1 and /healthz/v2.
       • Exact match: The URL can be accessed only when it is fully matched. For example, if the URL is set to /healthz, only /healthz can be accessed.
       • Regular expression: The URL is matched based on the regular expression. For example, if the regular expression is /[A-Za-z0-9_.-]+/test, all URLs that comply with this rule can be accessed, for example, /abcA9/test and /v1-Ab/test. Two regular expression standards are supported: POSIX and Perl.
     • URL: access path to be registered, for example, /healthz.
     • Destination Service: Select an existing Service or create a Service. Services that do not meet search criteria are automatically filtered out.
     • Destination Service Port: Select the access port of the destination Service.
     • Operation: Click Delete to delete the configuration.
   • Annotation: Ingresses provide some advanced CCE functions, which are implemented by annotations. When you use kubectl to create a container, annotations will be used. For details, see Creating an Ingress - Automatically Creating a Load Balancer and Creating an Ingress - Interconnecting with an Existing Load Balancer.

4. After the configuration is complete, click Create. After the ingress is created, it is displayed in the ingress list.

   On the ELB console, you can view the ELB automatically created through CCE. The default name is cce-lb-ingress.UID. Click the ELB name to access its details page. On the Listeners tab page, view the route settings of the ingress, including the URL, listener port, and backend server group port.

   

   After the ingress is created, upgrade and maintain the selected load balancer on the CCE console. Do not maintain the load balancer on the ELB console. Otherwise, the ingress service may be abnormal.

   Figure 1 ELB routing configuration
   

5. Access the /healthz interface of the workload, for example, workload defaultbackend.
   1. Obtain the access address of the /healthz interface of the workload. The access address consists of the load balancer IP address, external port, and mapping URL, for example, 10.**.**.**:80/healthz.
   2. Enter the URL of the /healthz interface, for example, http://10.**.**.**:80/healthz, in the address box of the browser to access the workload, as shown in Figure 2.
      Figure 2 Accessing the /healthz interface of defaultbackend
      

Related Operations

The Kubernetes ingress structure does not contain the property field. Therefore, the ingress created by the API called by client-go does not contain the property field. CCE provides a solution to ensure compatibility with the Kubernetes client-go. For details about the solution, see How Can I Achieve Compatibility Between Ingress's property and Kubernetes client-go?