fix(ci): pull-ci-redhat-developer-rhdh-main-e2e-osd-gcp-operator-nightly#4419
Conversation
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
|
The container image build workflow finished with status: |
|
The container image build workflow finished with status: |
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-ocp-helm |
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
1 similar comment
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
Fortune-Ndlovu
changed the title
Fortune-Ndlovu
force-pushed
the
RHDHBUGS-1136-e2e-OSD-GCP-with-Operator-deployment-not-starting-fix
branch
from
ec49f17 to
678ec25
Compare
|
/test e2e-osd-gcp-operator-nightly |
|
The container image build workflow finished with status: |
The OSD-GCP environment operates in a restricted network that cannot access GitHub Container Registry (ghcr.io). When the install-dynamic-plugins init container tried to download the Tekton plugin from ghcr.io, it timed out causing deployment failures. This fix disables the Tekton plugin for OSD-GCP deployments, matching the behavior of other K8s cloud providers (AKS, EKS, GKE) where Tekton is also not installed. Root cause: Network timeout accessing ghcr.io from OSD-GCP cluster Error: dial tcp 140.82.116.33:443: i/o timeout Fixes: RHDHBUGS-1136 Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
…oyment-not-starting-fix
|
/test e2e-osd-gcp-operator-nightly |
|
The container image build workflow finished with status: |
The plugin catalog index (dynamic-plugins.default.yaml) from quay.io/rhdh/plugin-catalog-index:1.10 contains many OCI plugins from ghcr.io. OSD-GCP clusters cannot access ghcr.io due to network restrictions, causing init container crashes when trying to inspect these plugins (Tekton, JFrog Artifactory, Jenkins, etc.). Previous fix (disabling Tekton only) was insufficient - the init container would just fail on the next ghcr.io plugin in sequence. This fix completely disables the catalog index by setting: global.dynamic.includes: [] Only pre-bundled plugins from the RHDH container image will be used on OSD-GCP, which are accessible from the local filesystem. Root cause: Network timeout accessing ghcr.io:443 from OSD-GCP cluster Errors seen on: backstage-community-plugin-jfrog-artifactory, backstage-community-plugin-tekton, and potentially all ghcr.io plugins Fixes: RHDHBUGS-1136 Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
|
/test e2e-osd-gcp-operator-nightly |
Previous approach (disabling catalog index) broke plugins that use
{{inherit}} because they rely on the catalog index for version info.
This fix keeps the catalog index but selectively disables ghcr.io
plugins that cannot be accessed from OSD-GCP due to network restrictions:
- backstage-community-plugin-github-issues
- roadiehq-backstage-plugin-github-pull-requests
- backstage-community-plugin-github-actions
- backstage-community-plugin-quay and quay-backend
- backstage-community-plugin-scaffolder-backend-module-quay
- immobiliarelabs-backstage-plugin-gitlab-backend
The catalog index remains active for other plugins that can inherit
versions from it.
Fixes: RHDHBUGS-1136
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
|
/test e2e-osd-gcp-operator-nightly |
… catalog index
OSD-GCP clusters cannot access ghcr.io (GitHub Container Registry) due to
network restrictions. The catalog index (quay.io/rhdh/plugin-catalog-index:1.10)
contains 21 ghcr.io plugins with explicit versions (bs_1.45.3__X.X.X format).
Previous attempts to disable these plugins using {{inherit}} failed because
the catalog index references them with explicit versions, not {{inherit}}.
The plugin installer treats package:version and package:{{inherit}} as
different package references.
Changes:
- Updated diff-values_showcase_OSD-GCP.yaml to disable all 21 ghcr.io plugins
with their exact versions from the catalog index
- Updated diff-values_showcase-rbac_OSD-GCP.yaml with the same plugin disables
- Updated docs/e2e-tests/CI.md to document all disabled plugins with versions
This ensures the init container (install-dynamic-plugins) won't attempt to
access ghcr.io for any plugins, preventing network timeouts and CrashLoopBackOff.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
|
/test e2e-osd-gcp-operator-nightly |
1 similar comment
|
/test e2e-osd-gcp-operator-nightly |
|
/test e2e-osd-gcp-operator-nightly |
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Fortune-Ndlovu
force-pushed
the
RHDHBUGS-1136-e2e-OSD-GCP-with-Operator-deployment-not-starting-fix
branch
from
41ee3e5 to
381d6ca
Compare
|
|
/test e2e-osd-gcp-operator-nightly |
|
@Fortune-Ndlovu: The following test failed, say
Full PR test history. Your PR dashboard. DetailsInstructions 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. |
1 participant
Description
fix(ci): pull-ci-redhat-developer-rhdh-main-e2e-osd-gcp-operator-nightly
PR acceptance criteria
Please make sure that the following steps are complete:
How to test changes / Special notes to the reviewer