View this page in the latest version of Appian. Exposing Appian Outside Kubernetes Share Share via LinkedIn Reddit Email Copy Link Print On This Page This page describes how to expose Appian sites outside of Kubernetes. Exposing the webapp or Apache Web Server (httpd) component allows Appian to send and receive web traffic. Any exposure should be used with appropriate security measures. Your organization is responsible for handling ingress securely. How ingress works In Kubernetes, an Ingress resource defines rules for routing external HTTP/https traffic into services inside a cluster. An Ingress controller is responsible for fulfilling the ingress. Appian sites in Kubernetes should work with a variety of Ingress controllers provided both the Ingress resource and controller are properly configured. Configure an Ingress resource To enable and configure an Ingress resource for your environment, make the following changes to the Appian custom resource (CR) configuration file: Set the .spec.ingress.enabled field to true. Set the .spec.ingress.annotations field as appropriate: The specific annotation you need depends on your ingress controller. If your ingress controller uses load balancers that run outside the cluster, configure an annotation that allows proper handling of non-2xx responses, such as HTTP 404. This configuration prevents health probes from incorrectly marking the httpd backend as unhealthy. Caution: If you do not configure this type of annotation, the external load balancer may mark the Appian httpd backend as unhealthy and block access to the site, even if Appian is running correctly. Note: If you've enabled network policies via the .spec.networkPolicies.enabled field on Appian custom resources, make sure to configure your network policies appropriately as well. Example The following configuration uses an AWS Application Load Balancer (ALB) annotation: 1 2 3 4 5 6 spec: ingress: enabled: true annotations: # AWS ALB: accept 200 and 404 as healthy responses alb.ingress.kubernetes.io/success-codes: "200,404" Tip: The exact syntax and accepted status code format can vary by controller version. Verify the correct values for your specific setup using your Ingress controller's documentation. Configure Gateway API routes Some ingress controllers, such as Envoy Gateway, do not process Ingress resources and instead implement the Kubernetes Gateway API. Starting with Appian operator v0.218.0, the operator can create HTTPRoute resources for these controllers instead of an Ingress resource. The operator only creates routes that attach to an existing Gateway. You are responsible for creating the Gateway, its listeners, and the TLS certificates those listeners serve. Every hostname Appian serves must be covered by a certificate on one of the Gateway's listeners. Prerequisites The operator only needs the Gateway API custom resource definitions if you use Gateway API routes; clusters that use Ingress resources do not need them. To use Gateway API routes, install the Gateway API custom resource definitions in the cluster before you install or upgrade the Appian operator's Helm chart. The chart only grants the operator permission to manage HTTPRoute resources if the cluster serves them, and the operator only checks whether they are served when it starts. If you install the Gateway API custom resource definitions after installing the operator, re-run helm upgrade for the operator and then restart the operator's controllers pod. Enable Gateway API routes To have the operator create routes for your environment, make the following changes to the Appian custom resource (CR) configuration file: Set the .spec.ingress.enabled field to false. Ingress and Gateway API routes cannot be enabled at the same time. Set the .spec.httpRoute.enabled field to true. Set the .spec.httpRoute.parentRefs field to the Gateway the routes should attach to. To attach to a Gateway in a different namespace, the Gateway's listeners must allow routes from the Appian custom resource's namespace. If you've enabled network policies, allow the Gateway's data plane pods to reach Appian by adding them to .spec.httpd.additionalNetworkPolicyIngressRules, or to .spec.webapp.additionalNetworkPolicyIngressRules if .spec.httpd is not defined. If RPA is enabled, also add them to .spec.rpa.additionalNetworkPolicyIngressRules, as RPA's paths are routed directly to RPA. Without these rules, the routes are accepted but requests are dropped without a response. Tip: If you don't name a listener with sectionName or port in .spec.httpRoute.parentRefs, the routes attach to every compatible listener on the Gateway, which may include plaintext HTTP listeners. The following configuration attaches Appian's routes to the https listener of a Gateway named eg in the envoy-gateway-system namespace, and allows that Gateway's pods through httpd's network policy: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 spec: ingress: enabled: false httpRoute: enabled: true parentRefs: - name: eg namespace: envoy-gateway-system sectionName: https httpd: additionalNetworkPolicyIngressRules: - from: - namespaceSelector: matchLabels: kubernetes.io/metadata.name: envoy-gateway-system podSelector: matchLabels: gateway.envoyproxy.io/owning-gateway-name: eg Differences from Ingress Rules in .spec.ingress.additionalIngressRules are not translated into routes. To serve additional paths, create your own HTTPRoute that attaches to the same Gateway and hostname. The operator only manages the routes it creates and leaves yours unchanged. Annotations in .spec.ingress.annotations are not copied to the routes. Use .spec.httpRoute.annotations for route annotations, and configure behavior such as health check status codes on the Gateway or your controller's policy resources instead. See HTTPRouteConfig for the full list of fields. Feedback Was this page helpful? SHARE FEEDBACK Loading...