Skip to content
This repository has been archived by the owner on Aug 16, 2023. It is now read-only.

Latest commit

 

History

History
126 lines (75 loc) · 5.39 KB

CHANGELOG.rst

File metadata and controls

126 lines (75 loc) · 5.39 KB
Raw

What has changed in ci-artifacts?

(Organized release by release)

Changes since version 0.1 (June 2021)

Toolbox

  • Add nfd test_master_branch protocol #179
    • new toolbox command: toolbox/nfd-operator/deploy_from_commit.sh <git repository> <git reference> to deploy NFD Operator from a custom commit.
  • Support for running NTO e2e tests #185
  • new toolbox command: toolbox/nto/run_e2e_test.sh <git repository> <git reference> to run the NTO e2e testsuite from a given commit.
  • Add --base_machineset flag to run_toolbox.py cluster set_scale #243
    • new flag to set the base machineset, from which the new machineset will be derived.

Retro-compatibility breaks

Bug fixes

  • Use subscriptions.operators.coreos.com instead of subscriptions to avoid conflicts with Knative subscriptions #207 #208

Features of version 0.1 (June 2021)

  • toolbox/nfd/deploy_from_operatorhub.sh was moved to toolbox/nfd-operator/deploy_from_operatorhub.sh

Bug fixes

  • toolbox/local-ci/deploy.sh <ci command> <git repository> <git reference> was fixed #179

Other changes

  • Introduce a Github Action for checking ansible variable consistency #196
    • toolbox/repo/validate_role_files.py is a new script to ensure that all the Ansible variables defining a filepath (roles/) do point to an existing file
    • toolbox/repo/validate_role_vars_used.py is a new script to ensure that all the Ansible variables defined are actually used in their role (with an exception for symlinks)
  • gpu_operator_deploy_from_operatorhub: allow overriding subscription.spec.installPlanApproval `#219<https://github.com/openshift-psap/ci-artifacts/pull/219>`_
    • ./toolbox/gpu-operator/deploy_from_operatorhub.sh can receive a new flag -install-plan=Manual|Automatic (Manual is the default) to override the Subscription install-plan approval setting when deploying from OperatorHub.

Changes since version May 7th, 2021

Retro-compatibility breaks

  • Change scaleup to set_scale - supported scale other than just 1 node #139
    • toolbox/cluster/scaleup.sh has been removed,
    • toolbox/cluster/set_scale.sh has been introduced as a replacement.
  • Add easy ways to test the entitlement #120
    • toolbox/entitlement/deploy.sh --machine-configs /path/to/machineconfigs has been removed
    • toolbox/entitlement/deploy.sh --pem /path/to/key.pem should be used instead. See there for a function to extract the PEM key from a machine-configs resource file.
  • toolbox: rename entitlement/test.sh -> entitlement/test_cluster.sh #166
    • toolbox/entitlement/test.sh was renamed into
    • toolbox/entitlement/test_cluster.sh

Other changes

  • Change scaleup to set_scale - supported scale other than just 1 node #139

    • toolbox/cluster/set_scale.sh has been introduced to control the scale (node count, of a given AWS instance-type) of a cluster

  • Add easy ways to test the entitlement #120

    • new commands to test a PEM key before deploying it:
      • toolbox/entitlement/test_in_podman.sh /path/to/key.pem
      • toolbox/entitlement/test_in_cluster.sh /path/to/key.pem

  • gpu_operator_set_repo-config: new role to set spec.driver.repoConfig #124

    • new option to deploy a custom PEM CA file, to access private repo mirrors
      • toolbox/entitlement/deploy.sh --pem </path/to/key.pem> [--ca </path/to/key.ca.pem>]
    • new command to configure the GPU Operator with a given repo-list file
      • toolbox/gpu-operator/set_repo-config.sh <path/to/repo.list> [<dest-dir>]

  • gpu_operator_deploy_from_operatorhub: add support for setting the channel # <#173>

    • toolbox/gpu-operator/deploy_from_operatorhub.sh [<version> [<channel>]]

CI Image and Testing

  • gpu_operator_set_repo-config: new role to set spec.driver.repoConfig #124

Bug fixes