Feature/operator version bump#372
Conversation
simonpriadka
force-pushed
the
feature/operator-version-bump
branch
from
9365c45 to
84cfb3f
Compare
simonpriadka
force-pushed
the
feature/operator-version-bump
branch
from
84cfb3f to
d4da358
Compare
| | Red Hat AMQ Broker | 7.14.0-opr-1 | 7.14.x | https://github.com/rh-messaging/activemq-artemis-operator | | As available on the OpenShift OperatorHub | | ||
| | Infinispan | 2.4.18 | stable | https://github.com/infinispan/infinispan-operator | | As available on the OpenShift OperatorHub (community-operators) | | ||
| | Red Hat DataGrid | 8.5.12 | stable | https://github.com/infinispan/infinispan-operator | | As available on the OpenShift OperatorHub | | ||
| | Red Hat DataGrid | 8.6.3 | stable | https://github.com/infinispan/infinispan-operator | | As available on the OpenShift OperatorHub | |
There was a problem hiding this comment.
I see
in current cluster ... have you considered upgrading to this operator version? what Data Grid version would it deploy? I suppose:
| | Red Hat DataGrid | 8.6.3 | stable | https://github.com/infinispan/infinispan-operator | | As available on the OpenShift OperatorHub | | ||
| | Kafka provided by Strimzi | 0.48.0 | stable | https://github.com/strimzi/strimzi-kafka-operator | | As available on the OpenShift OperatorHub (community-operators) | | ||
| | Streams for Apache Kafka | 3.0.1-3 | stable | https://github.com/strimzi/strimzi-kafka-operator | | As available on the OpenShift OperatorHub | | ||
| | Streams for Apache Kafka | 3.2.0-24 | stable | https://github.com/strimzi/strimzi-kafka-operator | | As available on the OpenShift OperatorHub | |
There was a problem hiding this comment.
I see this in the cluster we are currently using:
have you considered upgrading to this operator version?
| | Streams for Apache Kafka | 3.2.0-24 | stable | https://github.com/strimzi/strimzi-kafka-operator | | As available on the OpenShift OperatorHub | | ||
| | Keycloak | 26.4.4 | fast | https://github.com/keycloak/keycloak/tree/main/operator | | Latest Keycloak, based on Quarkus. | | ||
| | Red Hat Build of keycloak (RHBK) | 26.4.2-opr.1 | stable-v26.4 | https://github.com/keycloak/keycloak/tree/main/operator | | Latest Keycloak, based on Quarkus. | | ||
| | Red Hat Build of keycloak (RHBK) | 26.6.3-opr.1 | stable-v26.6 | https://github.com/keycloak/keycloak/tree/main/operator | | Latest Keycloak, based on Quarkus. | |
There was a problem hiding this comment.
I see this in the cluster we are currently using:
have you considered upgrading to this operator version?
| import org.keycloak.v1alpha1.KeycloakClient; | ||
| import org.keycloak.v1alpha1.KeycloakRealm; | ||
| import org.keycloak.v1alpha1.KeycloakUser; | ||
| import org.jboss.intersmash.rhsso.v1alpha1.Keycloak; |
There was a problem hiding this comment.
why did we change the package for the CRDs? is it just for Keycloak?
There was a problem hiding this comment.
yep, to separate rh sso from Keycloak
| // .failFast(ffCheck) | ||
| // .level(Level.DEBUG) | ||
| // .waitFor(); | ||
| if (replicas == 0) { |
There was a problem hiding this comment.
isn't this assuming we always go from replicas>0 to replicas=0 ? what if the cluster starts with 0 replicas? would it make sense to test ready.size() == 0 in io.amq.broker.v1beta1.activemqartemisstatus.PodStatus ?
There was a problem hiding this comment.
I've been observing this and the behavior when scaling down to 0 is, that the pod status will contain stopped so ready.size() == 0 won't work, as getReady() will be null in this case.
| <!-- | ||
| Separate RHSSO 7 (keycloak.org) CRDs from modern RHBK (k8s.keycloak.org) CRDs | ||
| --> | ||
| <packageOverrides> |
There was a problem hiding this comment.
RH SSO uses v1alpha1 whilst upgraded Keycloak moved to v2beta1. This was purely for the provisioners to take the distinctive imports
|
@simonpriadka thank you for your PR! I added a few questions |
2 participants
Description
This PR updates operator versions and CRDs for EAP 8.2, EAP XP 6, and EAP XP 7 profiles, along with significant refactoring of operator provisioning and CRD organization.
Key Changes:
Operator version bumps:
amqstreamstoamq-streams)CRD reorganization: Separated operator CRDs into dedicated subdirectories (
provisioners/src/main/resources/crds/default/{activemq,infinispan,keycloak,rhsso}) for better organization and clarityRH SSO separation: Separated Red Hat SSO (legacy) from Keycloak (modern) with dedicated CRDs and models
Bug fixes:
scale()method inActiveMQOperatorProvisionerOperatorProvisionerimplementation issuesDocumentation updates: Updated
docs/Operator-Based-Provisioning.mdto reflect new operator versionsType of change
Checklist
CI Runs: eap-8.x-cross-product-face#427