OCPBUGS-41892: Single Watch on GWAPI CRD

[Thealisyed]

A single CRD watch was added with a predicate to filter CRDs based on the gatewayapiv1.GroupName. This ensures that only relevant CRDs are watched, reducing unnecessary logging when the Gateway API feature is not enabled.

[openshift-ci-robot]

@Thealisyed: This pull request references Jira Issue OCPBUGS-41892, which is invalid:

• expected the bug to target the "4.18.0" version, but no target version was set

Comment /jira refresh to re-evaluate validity if changes to the Jira bug are made, or edit the title of this pull request to link to a different bug.

The bug has been updated to refer to the pull request using the external bug tracker.

In response to this:

A single CRD watch was added with a predicate to filter CRDs based on the gatewayapiv1.GroupName. This ensures that only relevant CRDs are watched, reducing unnecessary logging when the Gateway API feature is not enabled.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository.

[Thealisyed]

/jira refresh

[openshift-ci-robot]

@Thealisyed: This pull request references Jira Issue OCPBUGS-41892, which is valid. The bug has been moved to the POST state.

3 validation(s) were run on this bug

• bug is open, matching expected state (open)
• bug target version (4.18.0) matches configured target version for branch (4.18.0)
• bug is in the state ASSIGNED, which is one of the valid states (NEW, ASSIGNED, POST)

Requesting review from QA contact:
/cc @lihongan

In response to this:

/jira refresh

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository.

[Thealisyed]

/test e2e-aws-ovn-single-node

[Thealisyed]

/retest-required

[Miciah]

(Whoops, I've had this review pending, waiting for me to submit it, for 5 days...)

Is it true that the existing watch only generates reconciliation requests when the Gateway API feature is not enabled?

In any case, please add a little more detail in the commit message. For example:

gatewayapi: Use single watch for CRDs

Replace the multiple watches for Gateway API CRDs with a single CRD watch, with
a predicate to filter CRDs based on the gatewayapiv1.GroupName.  This ensures
that only relevant CRDs are watched, reducing unnecessary reconciliation
requests and log messages.

This commit fixes OCPBUGS-41892.

https://issues.redhat.com/browse/OCPBUGS-41892

[candita]

/assign
/assign @Miciah

[candita]

@Thealisyed please investigate the failures in the CI tests, especially e2e-aws-gatewayapi, which is the GWAPI test. I ran a retest to see if the issue with imagestreams had gone away.

[candita]

/retest

[candita]

@Thealisyed please note you will need to squash your three commits before merge.

[candita]

Process did not finish before 4h0m0s timeout

/test e2e-aws-ovn-serial

[Miciah]

/approve

The changes look fine to me. I did have a suggestion for the commit message: #1165 (review). We generally follow the upstream conventions: https://www.kubernetes.dev/docs/guide/pull-requests/#commit-message-guidelines

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: Miciah

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [Miciah]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[candita]

/test e2e-aws-ovn-techpreview

[Thealisyed]

@Miciah Apologies for the confusion! I initially misunderstood your comment and thought you wanted a longer commit message. I now see you're suggesting adding more detail, like in the example you shared. I'll update the commit message to follow the upstream guidelines. Thanks for the clarification!

[Thealisyed]

/retest-required

[Thealisyed]

/retest-required

[Thealisyed]

/test e2e-aws-operator

[candita]

@Thealisyed - before you test again, do a little analysis on the failures, and write them up like this. Then the next time they fail, if this failure keeps happening, you'll need to figure or get help to figure out why.

Two unit test failures:

TestAll/serial/TestRouterCompressionOperation
router_compression_test.go:245: failed to create service openshift-ingress/hello-pod-bpcnc: Internal error occurred: failed to allocate a serviceIP: rpc error: code = Unavailable desc = error reading from server: read tcp 10.0.89.66:43566->10.0.18.210:2379: read: connection reset by peer

TestAll/serial/TestCanaryWithMTLS
canary_test.go:343: Polling for canary success condition...
canary_test.go:346: Failed to get IngressController: Get "https://api.ci-op-ki1llnl8-9e7c5.origin-ci-int-aws.dev.rhcloud.com:6443/apis/operator.openshift.io/v1/namespaces/openshift-ingress-operator/ingresscontrollers/default": http2: client connection lost
canary_test.go:361: Polling terminated with unexpected error: Get "https://api.ci-op-ki1llnl8-9e7c5.origin-ci-int-aws.dev.rhcloud.com:6443/apis/operator.openshift.io/v1/namespaces/openshift-ingress-operator/ingresscontrollers/default": http2: client connection lost

/test e2e-aws-operator-techpreview

[Thealisyed]

"Operator "console-operator\ + cluster-samples-operator" produces more watch requests than expected [for both]
seems unrelated to changes will test again and observe

/retest

[openshift-ci]

@Thealisyed: all tests passed!

Full PR test history. Your PR dashboard.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here.

[candita]

@Thealisyed, in the e2e-aws-gatewayapi test logs, where the featuregate is on, I'm seeing some errors that should be investigated: https://gcsweb-ci.apps.ci.l2s4.p1.openshiftapps.com/gcs/test-platform-results/pr-logs/pull/openshift_cluster-ingress-operator/1165/pull-ci-openshift-cluster-ingress-operator-master-e2e-aws-gatewayapi/1862083335728140288/artifacts/e2e-aws-gatewayapi/gather-extra/artifacts/pods/openshift-ingress-operator_ingress-operator-58d85999bb-l6cpv_ingress-operator.log

Can you check if this happens on a cluster without your changes? If so, please open a new bug that points to the logs. If not, we need to find out what's in this PR that would cause the errors.

Specifically, lots of these:

2024-11-28T11:51:57.292Z ERROR operator.init controller/controller.go:263 Reconciler error {"controller": "gatewayclass_controller", "object": {"name":"openshift-default"}, "namespace": "", "name": "openshift-default", "reconcileID": "e18f9589-f527-4802-87aa-bc6ebb8240fd", "error": "failed to create ServiceMeshControlPlane openshift-ingress/openshift-gateway: Internal error occurred: failed calling webhook "smcp.mutation.maistra.io": failed to call webhook: Post "https://istio-operator-service.openshift-operators.svc:443/mutate-smcp?timeout=10s\": no endpoints available for service "istio-operator-service"", "errorCauses": [{"error": "failed to create ServiceMeshControlPlane openshift-ingress/openshift-gateway: Internal error occurred: failed calling webhook "smcp.mutation.maistra.io": failed to call webhook: Post "https://istio-operator-service.openshift-operators.svc:443/mutate-smcp?timeout=10s": no endpoints available for service "istio-operator-service""}]}

and then eventually it is able to create the servicemeshcontrolplane:

2024-11-28T11:52:17.801Z INFO operator.gatewayclass_controller gatewayclass/servicemeshcontrolplane.go:46 created ServiceMeshControlPlane {"namespace": "openshift-ingress", "name": "openshift-gateway"}

[candita]

Hold for a check on logs
/hold
Otherwise
/lgtm

[rhamini3]

Pre merge verified, Gateway logs does not flood the cluster-ingress-operator logs

% oc logs ingress-operator-747dd58b69-xwpsk -n openshift-ingress-operator | grep gateway
Defaulted container "ingress-operator" out of: ingress-operator, kube-rbac-proxy
2024-12-18T17:00:21.604Z	INFO	operator.init	controller/controller.go:231	Starting EventSource	{"controller": "gatewayapi_controller", "source": "kind source: *v1.FeatureGate"}
2024-12-18T17:00:21.605Z	INFO	operator.init	controller/controller.go:231	Starting EventSource	{"controller": "gatewayapi_controller", "source": "kind source: *v1.CustomResourceDefinition"}
2024-12-18T17:00:21.605Z	INFO	operator.init	controller/controller.go:231	Starting Controller	{"controller": "gatewayapi_controller"}
2024-12-18T17:00:22.130Z	INFO	operator.init	controller/controller.go:231	Starting workers	{"controller": "gatewayapi_controller", "worker count": 1}
2024-12-18T17:00:22.130Z	INFO	operator.gatewayapi_controller	controller/controller.go:116	reconciling	{"request": {"name":"cluster"}}

/label qe-approved

[openshift-ci-robot]

@Thealisyed: This pull request references Jira Issue OCPBUGS-41892, which is invalid:

• expected the bug to target either version "4.19." or "openshift-4.19.", but it targets "4.18.0" instead

Comment /jira refresh to re-evaluate validity if changes to the Jira bug are made, or edit the title of this pull request to link to a different bug.

The bug has been updated to refer to the pull request using the external bug tracker.

In response to this:

A single CRD watch was added with a predicate to filter CRDs based on the gatewayapiv1.GroupName. This ensures that only relevant CRDs are watched, reducing unnecessary logging when the Gateway API feature is not enabled.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository.

[Miciah]

/jira refresh

[openshift-ci-robot]

@Miciah: This pull request references Jira Issue OCPBUGS-41892, which is valid.

3 validation(s) were run on this bug

• bug is open, matching expected state (open)
• bug target version (4.19.0) matches configured target version for branch (4.19.0)
• bug is in the state POST, which is one of the valid states (NEW, ASSIGNED, POST)

No GitHub users were found matching the public email listed for the QA contact in Jira ([email protected]), skipping review request.

In response to this:

/jira refresh

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository.