Installing Knative Eventing using YAML files
Before installing Knative, you must meet the following prerequisites:
For prototyping purposes, Knative works on most local deployments of Kubernetes. For example, you can use a local, one-node cluster that has 3 CPUs and 4 GB of memory.
Tip
You can install a local distribution of Knative for development purposes using the
For production purposes, it is recommended that:
- If you have only one node in your cluster, you need at least 6 CPUs, 6 GB of memory, and 30 GB of disk storage.
- If you have multiple nodes in your cluster, for each node you need at least 2 CPUs, 4 GB of memory, and 20 GB of disk storage.
- You have a cluster that uses Kubernetes v1.22 or newer.
- You have installed the kubectl CLI.
- Your Kubernetes cluster must have access to the internet, because Kubernetes needs to be able to fetch images. To pull from a private registry, see .
Caution
The system requirements provided are recommendations only. The requirements for your installation might vary, depending on whether you use optional components, such as a networking layer.
Install Knative Eventing
To install Knative Eventing:
Install the required custom resource definitions (CRDs) by running the command:
Install the core components of Eventing by running the command:
kubectl apply -f https://github.com/knative/eventing/releases/download/knative-v1.7.5/eventing-core.yaml
Info
For information about the YAML files in Knative Eventing, see .
Success
Monitor the Knative components until all of the components show a STATUS
of Running
or Completed
. You can do this by running the following command and inspecting the output:
kubectl get pods -n knative-eventing
Example output:
NAME READY STATUS RESTARTS AGE
eventing-controller-7995d654c7-qg895 1/1 Running 0 2m18s
eventing-webhook-fff97b47c-8hmt8 1/1 Running 0 2m17s
Optional: Install a default Channel (messaging) layer
The following tabs expand to show instructions for installing a default Channel layer. Follow the procedure for the Channel of your choice:
Apache Kafka ChannelIn-Memory (standalone)NATS Channel
The following commands install the KafkaChannel and run event routing in a system namespace. The knative-eventing
namespace is used by default.
Install the Kafka controller by running the following command:
kubectl apply -f https://github.com/knative-sandbox/eventing-kafka-broker/releases/download/knative-v1.7.6/eventing-kafka-controller.yaml
Install the KafkaChannel data plane by running the following command:
kubectl apply -f https://github.com/knative-sandbox/eventing-kafka-broker/releases/download/knative-v1.7.6/eventing-kafka-channel.yaml
-
kubectl apply -f https://github.com/knative-sandbox/eventing-kafka-broker/releases/download/knative-v1.7.6/eventing-kafka-post-install.yaml
Warning
This simple standalone implementation runs in-memory and is not suitable for production use cases.
Install an in-memory implementation of Channel by running the command:
kubectl apply -f https://github.com/knative/eventing/releases/download/knative-v1.7.5/in-memory-channel.yaml
Install the NATS Streaming Channel by running the command:
You can change the default channel implementation by following the instructions described in the section.
The following tabs expand to show instructions for installing the Broker layer. Follow the procedure for the Broker of your choice:
Apache Kafka BrokerMT-Channel-basedRabbitMQ Broker
The following commands install the Apache Kafka Broker and run event routing in a system namespace. The knative-eventing
namespace is used by default.
Install the Kafka controller by running the following command:
kubectl apply -f https://github.com/knative-sandbox/eventing-kafka-broker/releases/download/knative-v1.7.6/eventing-kafka-controller.yaml
Install the Kafka Broker data plane by running the following command:
kubectl apply -f https://github.com/knative-sandbox/eventing-kafka-broker/releases/download/knative-v1.7.6/eventing-kafka-broker.yaml
For more information, see the documentation.
This implementation of Broker uses Channels and runs event routing components in a system namespace, providing a smaller and simpler installation.
Install this implementation of Broker by running the command:
kubectl apply -f https://github.com/knative/eventing/releases/download/knative-v1.7.5/mt-channel-broker.yaml
To customize which Broker Channel implementation is used, update the following ConfigMap to specify which configurations are used for which namespaces:
apiVersion: v1
kind: ConfigMap
metadata:
name: config-br-defaults
namespace: knative-eventing
data:
default-br-config: |
clusterDefault:
brokerClass: MTChannelBasedBroker
apiVersion: v1
kind: ConfigMap
name: imc-channel
namespace: knative-eventing
# This allows you to specify different defaults per-namespace,
# in this case the "some-namespace" namespace will use the Kafka
# channel ConfigMap by default (only for example, you will need
# to install kafka also to make use of this).
namespaceDefaults:
some-namespace:
brokerClass: MTChannelBasedBroker
apiVersion: v1
kind: ConfigMap
name: kafka-channel
namespace: knative-eventing
The referenced
imc-channel
andkafka-channel
example ConfigMaps would look like:apiVersion: v1
kind: ConfigMap
metadata:
name: imc-channel
namespace: knative-eventing
data:
apiVersion: messaging.knative.dev/v1
kind: InMemoryChannel
---
apiVersion: v1
kind: ConfigMap
metadata:
name: kafka-channel
namespace: knative-eventing
data:
channel-template-spec: |
apiVersion: messaging.knative.dev/v1alpha1
kind: KafkaChannel
spec:
numPartitions: 3
replicationFactor: 1
Warning
In order to use the KafkaChannel, ensure that it is installed on your cluster, as mentioned previously in this topic.
- Install the RabbitMQ Broker by following the instructions in the RabbitMQ Knative Eventing Broker README.
For more information, see the in GitHub.
Install optional Eventing extensions
Apache Kafka SinkSugar ControllerGitHub SourceApache Kafka SourceApache CouchDB SourceVMware Sources and Bindings
Install the Kafka controller by running the command:
kubectl apply -f https://github.com/knative-sandbox/eventing-kafka-broker/releases/download/knative-v1.7.6/eventing-kafka-controller.yaml
Install the Kafka Sink data plane by running the command:
For more information, see the documentation.
Install the Eventing Sugar Controller by running the command:
kubectl apply -f https://github.com/knative/eventing/releases/download/knative-v1.7.5/eventing-sugar-controller.yaml
The Knative Eventing Sugar Controller reacts to special labels and annotations and produce Eventing resources. For example:
- When a namespace is labeled with
eventing.knative.dev/injection=enabled
, the controller creates a default Broker in that namespace. - When a Trigger is annotated with
eventing.knative.dev/injection=enabled
, the controller creates a Broker named by that Trigger in the Trigger’s namespace.
Enable the default Broker on a namespace (here
default
) by running the command:kubectl label namespace <namespace-name> eventing.knative.dev/injection=enabled
Where
<namespace-name>
is the name of the namespace.
A single-tenant GitHub source creates one Knative service per GitHub source.
A multi-tenant GitHub source only creates one Knative Service, which handles all GitHub sources in the cluster. This source does not support logging or tracing configuration.
To install a single-tenant GitHub source run the command:
kubectl apply -f https://github.com/knative-sandbox/eventing-github/releases/download/knative-v1.7.0/github.yaml
To install a multi-tenant GitHub source run the command:
kubectl apply -f https://github.com/knative-sandbox/eventing-github/releases/download/knative-v1.7.0/mt-github.yaml
To learn more, try the GitHub source sample
Install the Apache Kafka Source by running the command:
kubectl apply -f https://github.com/knative-sandbox/eventing-kafka-broker/releases/download/knative-v1.7.6/eventing-kafka-source.yaml
If you’re upgrading from the previous version, run the following command:
To learn more, try the .
Install the Apache CouchDB Source by running the command:
To learn more, read the Apache CouchDB source documentation.
Install VMware Sources and Bindings by running the command:
To learn more, try the .