Skip to content

Latest commit

 

History

History
67 lines (48 loc) · 3.79 KB

es-recipe.md

File metadata and controls

67 lines (48 loc) · 3.79 KB

Deploy Cloud Pak for Integration - IBM Event Streams capability

Overview

This IBM Event Streams recipe should provide a highly available deployment of IBM Event Streams on a Red Hat OpenShift Kubernetes Service on IBM Cloud as shown below.

es-qs

Infrastructure - kustomization.yaml

  1. Edit the Infrastructure layer ${GITOPS_PROFILE}/1-infra/kustomization.yaml and un-comment the following: 
    - argocd/namespace-ibm-common-services.yaml
- argocd/namespace-tools.yaml

Services - kustomization.yaml

  1. Edit the Services layer ${GITOPS_PROFILE}/2-services/kustomization.yaml uncomment the following: 
    - argocd/operators/ibm-eventstreams-operator.yaml
- argocd/instances/ibm-eventstreams-instance.yaml
- argocd/operators/ibm-foundations.yaml
- argocd/instances/ibm-foundational-services-instance.yaml
- argocd/operators/ibm-catalogs.yaml

Storage - ibm-eventstreams-instance.yaml

Make sure the storageClassName specified in ${GITOPS_PROFILE}/2-services/argocd/instances/ibm-eventstreams-instance.yaml suits your needs. By default, the storage class used for this recipe, which should be the storage class requested unless the use case or specific requirements dictate otherwise, is managed-nfs-storage since the IBM Technology Zone environments this recipe is meant for mount NFS storage. If you are using this recipe and the GitOps Framework that executes this recipe on another environment, make sure you select the appropriate storage class. Bear in mind that IBM Event Streams requires block storage.

High Availability - ibm-eventstreams-instance.yaml

Make sure the high availability specified in ${GITOPS_PROFILE}/2-services/argocd/instances/ibm-eventstreams-instance.yaml for each of the IBM Event Streams components suit your needs.

  • The Kafka and Zookeeper components are already set up highly available as you can see within spec.strimziOverrides.kafka.replicas and spec.strimziOverrides.zookeeper.replicas respectively.
  • The IBM Event Streams UI, the IBM Event Streams Apicurio Registry, the IBM Event Streams Admin API, the IBM Event Streams REST Producer and the IBM Event Streams Metrics Collector components can also be set up for high availability in their respective spec.adminUI, spec.apicurioRegistry, spec.adminApi, spec.restProducer and spec.collector sections by specifying the number of replicas for those too. By default, since no specific number of replicas is specified, only one pod for each of the aforementioned components will be deployed.

Security - ibm-eventstreams-instance.yaml

Make sure the security mechanisms to access Kafka suit your needs. IBM Event Streams provides different listeners within the spec.strimziOverrides.kafka.listeners:

  1. One external which will be SSL secured (external)
  2. Two internal listeners, one SSL secured (tls) and one non-SSL secured (plain)

All of these listeners can get configured for authentication, either SCRAM (scram-sha-512) or mTLS (tls). For more information, refer to the Kafka access section in the official IBM Event Streams documentation here.

Validation

After some time, you can validate your new instance of IBM Event Streams is ready by executing:

oc get eventstreams -n tools

Expected output is

NAME             STATUS
es-inst          Ready

You can now access your IBM Event Streams

oc get EventStreams es-inst -n tools -o=jsonpath='{.status.endpoints[?(@.name=="ui")].uri}'

The credentials for logging into the IBM Event Streams are admin/<password> where password is stored in a secret.

oc get secret platform-auth-idp-credentials -n ibm-common-services -o=jsonpath='{.data.admin_password}' | base64 -D