diff --git a/docs/versioned/admin/admin-overview.md b/docs/versioned/admin/admin-overview.md index 16f9f94133..29f8e546f6 100644 --- a/docs/versioned/admin/admin-overview.md +++ b/docs/versioned/admin/admin-overview.md @@ -9,38 +9,79 @@ function: explanation This page provides guidance for administrators on how to manage Knative on an existing Kubernetes cluster. + + +```mermaid +--- +config: + theme: mc + layout: elk + look: classic +--- +flowchart LR + subgraph Knative["Knative"] + direction LR + Serving["Serving"] + Eventing["Eventing"] + end + subgraph Plugins["Plugins"] + direction LR + net-istio["Istio"] + net-contour["Contour"] + net-gateway-api["Gateway API"] + event-kafka["Kafka"] + event-rabbitmq["RabbitMQ"] + event-nats["NATS"] + end + Dev(["Developers"]) --> dev-acts["Manages defined Kubernetes objects"] + dev-acts --> Serving & Eventing + Admin(["Administrators"]) --> admin-acts["Installs and configures"] + admin-acts --> Knative & Plugins + Serving --> net-impl["implements"] + net-impl --> net-istio & net-contour & net-gateway-api + Eventing --> event-impl["implements"] + event-impl --> event-kafka & event-rabbitmq & event-nats + dev-acts@{ shape: text} + admin-acts@{ shape: text} + net-impl@{ shape: text} + event-impl@{ shape: text} + style Serving fill:#757575,color:#FFFFFF + style net-istio fill:#757575,color:#FFFFFF + style net-contour fill:#757575,color:#FFFFFF + style net-gateway-api fill:#757575,color:#FFFFFF + style Dev fill:#FF6D00,color:#FFFFFF + style Admin fill:#2962FF,color:#FFFFFF + + +``` + As a cluster administrator, your responsibilities include managing the Kubernetes environment, installing cluster-wide components, and enabling developers to deploy applications on the cluster. Knative aims to simplify developer tasks, while aligning with existing management tools and processes. Knative includes a plugin system to integrate with existing infrastructure in the cluster, enabling Knative resources such as Routes and Brokers to be implemented using one of multiple underlying suppliers. For example, a Knative Eventing app can deliver events to a Broker that triggers a function based on the received event. In a testing cluster, the delivery might use an in-memory option, while a staging or production environment might use a cloud-provided Kafka service. Of particular interest to cluster administrators is that Knative supports customizable _default values_ on the parameters defined in resource YAML files. These configurations reduce the amount of environment configuration tasks developers needs to consider. -## Knative installations +## Installation decisions See the [Installation roadmap](../install/README.md#installation-roadmap) for prerequisites and installation steps. Your first installation decision is whether to use a YAML-based installation or use the Knative Operator. The Knative Operator is a custom controller that extends the Kubernetes API to install Knative components. If you just need to get acquainted with Knative at this time, you can install the [quickstart](../getting-started/quickstart-install.md). -## Configuring Knative +Knative installations are not permanent and you can install clusters differently depending on the situation. Although switching services and applications in real time may be possible, a best practice is to start with a new cluster. -Knative uses Kubernetes YAML manifests to define and configure system components. These manifests include core resources, custom resource definitions (CRDs), and extensibility features. As with Kubernetes, these configuration resources are declarative and can be managed using the `kubectl` CLI tool or with continuous delivery tools. - -### Resource scoping and namespaces +### Upgrades -Knative resources are namespaced. Knative adheres to the Kubernetes model of namespace-based isolation that lets you manage development teams and resources by assigning them to namespaces. +Administrators are generally responsible for performing upgrades to cluster infrastructure, apps, and services. Knative is designed and tested for continuous operation during upgrades and rollbacks, allowing you to: -Namespaces can also isolate boundaries for tooling such as logs, metrics, tracing, CI/CD integrations, and dashboards. The extent of this isolation depends on both the enforcement strategy and how consistently teams adhere to namespace boundaries. - -You can optimize and enforce isolation involving namespaces using standard Kubernetes mechanisms, including: - -- [Role-Based Access Control (RBAC)](https://kubernetes.io/docs/reference/access-authn-authz/rbac/) -- [Resource Quotas](https://kubernetes.io/docs/concepts/policy/resource-quotas/) -- [Network Policies](https://kubernetes.io/docs/concepts/services-networking/network-policies/) -- [Pod Security Standards](https://kubernetes.io/docs/concepts/security/pod-security-standards/) +- Upgrade or revert the Knative components while it is serving traffic, rather than needing a maintenance window. +- Downgrade by one Knative version. Downgrades work provided that no applications have used new features since the last upgrade. -### Configuring Knative components +## Configurations Knative configurations are performed by the following methods: -- Editing YAML manifests +- Editing YAML manifests and applying with the `kubectl` tool Modify resource definitions directly, including labels, annotations, and field values. You can use Kubernetes features such as [OPA](https://kubernetes.io/blog/2019/08/06/opa-gatekeeper-policy-and-governance-for-kubernetes/) and [Kyverno](https://kyverno.io) to enforce specific values on a resource type, or use ConfigMaps in plugin installations to set values at the cluster level. @@ -50,59 +91,61 @@ Knative configurations are performed by the following methods: - Using the Knative Operator - Some platform-wide settings can be managed declaratively using the Knative Operator. - -### Configuration tasks - -Knative documentation provides the following configuration procedures. This list is subject subject to change. - -Configurations for default settings: + Some platform-wide settings can be managed declaratively using the Knative Operator, installed with the `kn` Knative CLI plugin. You can manage the operator without using the `kn` CLI. The `kn` CLI manages only operator installations. -- [Broker defaults](../eventing/configuration/broker-configuration.md) -- [ConfigMap defaults](../serving/configuration/config-defaults.md) -- [Event source defaults](../eventing/configuration/sources-configuration.md) -- [Channel defaults](../eventing/configuration/channel-configuration.md) -- [Kafka channel defaults](../eventing/configuration/kafka-channel-configuration.md) -- [Domain names](../serving/using-a-custom-domain.md) -- [Ingress gateway replacement](../serving/setting-up-custom-ingress-gateway.md) -Configurations for new development: +For more information, see [Installing CLI tools](../client/README.md) -- [Deployment resources](../serving/configuration/deployment.md) -- [Istio access to deployed services](../serving/istio-authorization.md) -- [Namespace exclusion from webhooks](../serving/istio-authorization.md) +Knative uses Kubernetes YAML manifests to define and configure system components. These manifests include core resources, custom resource definitions (CRDs), and extensibility features. As with Kubernetes, these configuration resources are declarative and can be managed using the `kubectl` CLI tool or with continuous delivery tools. -Configurations for maintenance: +## Configuration summaries -- [Garbage collection](../serving/revisions/revision-admin-config-options.md) -- [High availability](../serving/config-ha.md) -- [Rollout duration for revisions](../serving/configuration/rolling-out-latest-revision-configmap.md) -- [Autoscaling of Kafka features](../eventing/configuration/keda-configuration.md) +The following sections provide an overview of the current configurations procedures of interest to Administrators. You make these configurations using `kubectl`, with some procedures applying YAML files. -Configurations for security encryptions: +### Networking -- [cert-manager](../serving/encryption/configure-certmanager-integration.md) -- [External domains](../serving/encryption/external-domain-tls.md) -- [Local domains](../serving/encryption/cluster-local-domain-tls.md) -- [system-internal](../serving/encryption/system-internal-tls.md) +| Configuration | ConfigMap | Description | +| -- | --- | --- | +| [Domain names](../serving/using-a-custom-domain.md) | `config-domain` | Configure and publish domains. | +| [Ingress gateway](../serving/setting-up-custom-ingress-gateway.md)| `config-istio` | For new clusters, you can configure your own gateway and underlying service. | +| [Istio authorization](../serving/istio-authorization.md) | NA | Grant authorization to your deployed Knative services. | -Configurations for extensions: +### Serving -- [Kafka Broker features](../serving/encryption/system-internal-tls.md) -- [Sugar Controller](../eventing/configuration/sugar-configuration.md) +| Configuration | ConfigMap | Description | +| -- | --- | --- | +| [Default configurations](../serving/configuration/config-defaults.md) | `config-defaults` | Default resource values such as performance, hardware, and storage settings. | +| [Deployment resources](../serving/configuration/deployment.md) | `config-deployment` | Kubernetes deployment resources that back Knative services. | +| [High-availability](../serving/config-ha.md) | NA | Configure ensure that APIs stay operational if a disruption occurs. | +| [Garbage collection](../serving/revisions/revision-admin-config-options.md) | `config-gc` | Disable and enable collection and set retention time values. | +| [Namespace exclusion from webhook](../serving/webhook-customizations.md) | NA | For performance concerns during an upgrade. | +| [Rollout duration for revisions](../serving/configuration/rolling-out-latest-revision-configmap.md) | `config-network` | Adjust rollout durations to accommodate longer request queues. | +| [Security - Certificates](../serving/encryption/configure-certmanager-integration.md) | NA | Describes how to manage automatic certificate provisioning. | +| [Security - Encryptions](../serving/encryption/encryption-overview.md) | `config-network` | Provides links to procedures for encrypting external domains, the local cluster, and system internal. | -Configurations for flagging features: +### Eventing -- [Serving features](../serving/configuration/feature-flags.md) -- [Eventing features](../eventing/features/README.md) +| Configuration | ConfigMap | Description | +| -- | --- | --- | +| [Broker defaults](../eventing/configuration/broker-configuration.md) | `config-br-defaults` | Specify your own broker class and channel, or use the default `MTChannelBasedBroker` Broker class and the ConfigMap of channel defaults. | +| [Broker features (Kafka)](../eventing/brokers/broker-types/kafka-broker/configuring-kafka-features.md) | `config-kafka-features` | Configure options for Broker interactions with Apache Kafka clusters. | +| [Channel defaults](../eventing/configuration/channel-configuration.md) | `default-ch-webhook` | Default configurations and labels to use for the channel. | +| [Channel defaults (Kafka)](../eventing/configuration/kafka-channel-configuration.md) | `kafka-channel` | Defines how KafkaChannel instances are created. Requires that KafkaChannel custom resource definitions (CRD) are installed.| +| [Event source defaults](../eventing/configuration/sources-configuration.md) |`config-ping-defaults` | Configure the PingSource default resources and the maximum data size for CloudEvents it produces. | +| [KEDA Autoscaling of Kafka Resources](../eventing/configuration/keda-configuration.md) |`config-kafka-features` | Configure how KEDA scales a KafkaSource, trigger, or subscription. Note: This feature is is Alpha pre-release. | +| [Sugar Controller](../eventing/sugar/README.md) |`config-sugar` | Configure the Sugar controller, which reacts to label configurations to produce or control eventing resources. See also [Knative Eventing Sugar Controller](../eventing/sugar/README.md). | -## Authorizations +## Securing Knative You can grant developers access to additional resources related to their namespace in other services, such as observability, logs, metrics, tracing, and dashboards. -## Upgrades +Knative resources are namespaced. Knative adheres to the Kubernetes model of namespace-based isolation that lets you manage development teams and resources by assigning them to namespaces. + +Namespaces can also isolate boundaries for tooling such as logs, metrics, tracing, CI/CD integrations, and dashboards. The extent of this isolation depends on both the enforcement strategy and how consistently teams adhere to namespace boundaries. -Administrators are generally responsible for performing upgrades cluster infrastructure and apps and services. Knative is designed and tested for continuous operation during upgrades and rollbacks, allowing you to: +You can optimize and enforce isolation involving namespaces using standard Kubernetes mechanisms, including: -- Upgrade or revert the Knative components while it is serving traffic, rather than needing a maintenance window. -- Downgrade one Knative version. Downgrades work provided that no applications have used new features since the last upgrade. +- [Role-Based Access Control (RBAC)](https://kubernetes.io/docs/reference/access-authn-authz/rbac/) +- [Resource Quotas](https://kubernetes.io/docs/concepts/policy/resource-quotas/) +- [Network Policies](https://kubernetes.io/docs/concepts/services-networking/network-policies/) +- [Pod Security Standards](https://kubernetes.io/docs/concepts/security/pod-security-standards/)