Kf release notes

Upgrade Tekton version to 1.3.2 and modify 'kf doctor' behavior and label selectors for pods to work with this Tekton version

Upgrade ASM to 1.27.1

Upgrade upload-pages-artifact dependency

Upgraded server-side dependencies - Tekton Pipelines, Config Connector

Changed version of php-buildpack to address build issue.

Upgraded server-side dependencies - Tekton Pipelines, ASM

Updated Go version used to build images and CLI tools

Upgraded server-side dependencies - Tekton Pipelines, ASM & Config Connector

Removed permissions on auditor roles that they could use to exec in containers if they used websockets with the Kubernetes API.

Added the ability to change the snapshot modes for Kaniko for faster snapshots and reduced disk I/O during builds.

Kf builds now show the number of files extracted from the uploaded source directory.

The default snapshot mode for Kaniko builds in the v2 buildpacks is now fast rather than robust.

The featureflag controller will no longer enqueue all namespaces when feature flags are changed, significantly reducing the load on the controller.

kf push now shows the amount of time source uploads took.

When spaces are updated, apps in the affected space will be enqueued on a best-effort basis to avoid blocking the main queue used for live interactions.

Added limits to containers.

Fixed set-env slowness.

Fixed issue that liveness probe is not set properly.

Fix Kf performance issue with high server load.

Removed NFS sidecar, built the logics into user-container.

Fixed the issue that App Pod will keep restarting when the mounted NFS volume is larger than cluster's ephemeral storage.

Fixed the issue that will cause delete-service stuck.

Kf will terminate Istio sidecars on completed tasks and builds freeing up IP addresses and compute resources for clusters running newer versions of Tekton.

kf doctor now works with GKE 1.22 clusters.

Kf will now install with GKE 1.22 clusters.

Builds will now be allowed up to 15 minutes to extract or publish images.

Removed kf migrate command.

Changed go version to 1.17.

Added support for mapping GID and UID for NFS.

Fixed issue that unbind-service will not retry when failed.

Fixed terminal size for Kf ssh.

Add Config Connector as a dependency of Kf.

Fixed bug in App controller, made App.Status.RouteConditions as an optional field.

Fixed bug in App controller, made App.Status.RouteConditions as an optional field.

Add a feature that supports adding node selectors for Kf Builds to isolate Kf Build pods in specific node pool.

Remove Config Connector as a dependency of Kf.

Added buildTimeout configuration feature.

Bug fix for Kf Build garbage collection.

Support for Anthos Service Mesh v1.12

Added buildDisableIstioSidecar configuration feature.

Added V3 kf-v2-to-v3-shim stack as build option.

Added buildPodResources configuration feature.

Added controllerCACerts configuration feature.

Added V3 Google stack as build option.

Added buildRetentionCount configuration feature.

Improved CLI performance.

Fixed an issue that could prevent SIGTERM from reaching an app.

Fixed an issue that caused extra reconciliation loops and logs.

Support for Anthos Service Mesh (ASM) v1.11+, which recommends ingress gateways be outside of the istio-system namespace.

Improved Workload Identity reliability.

Updated Tekton to v0.29.0.

Improved subresource API server resilience.

Updated Config Connect to v1.66.0.

Changed build ImagePullPolicy default from always download to prefer cached.

Resolved a scenario that could result in extra reconciliation loops and logs.

VCAP_APPLICATION route string no longer includes trailing slash.

Addressed a v2 buildpack condition that could prevent SIGTERM signals from propagating.

Anthos clusters on bare metal support promoted to GA.

Anthos clusters on VMware (GKE on-prem) support promoted to GA.

Added support for kubectl explain to inspect Kf CRDs .

New wrap-v2-buildpack experimentalcommand available.

Known issue:some buildpacks (ex: https://github.com/cloudfoundry/java-buildpack) produce directories that do not work with the pack CLI.

Fixed issue where kf doctor expects the ASM ingress gateway deployment to be in the kf namespace.

Addressed scenario where the Kf operator could overwrite ASM Gateway customization.

Fixed condition where CLI may not always show build logs.

Tekton dependency updated to v0.26.0.

Config Connector dependency updated to v1.60.

Removed downstream lifecycle dependency for v2 buildpacks that could result in kf push failing.

Fixed an error that can occur during the initialization of the subresource API.

Improved kf doctor reliability for Anthos on-prem clusters.

Kf for Anthos on-prem (Vsphere) is now available on the Public Previews page .

Addressed a potential panic in the kf build-logs command.

Changed flag and manifest validation for route and task fields so manifest routes are ignored when creating tasks.

When deleting resources, Kf will now delete any sub resources before returning success on the deletion request.

Kf CLI can now override the manifest.yaml app name.

New Early Access program (currently private preview).

New features may now start in preview and GA with the next Kf release.

[PREVIEW] NFS broker automatically configures NFS mounts on your cluster for Apps to bind to.

[PREVIEW] Schedule Tasks to run at recurring intervals specified using the unix-cron format.

[PREVIEW] Support for Anthos clusters on VMware via the Early Access program.

Automated assessment and migration of your existing CF foundation.

Minor reliability improvements.

Resolved race condition that could cause the first brokered service instance to fail.

Support for ASM 1.10.

Kf Cloud Service Broker for Google Cloud for Google managed services.

Allow customization of external-gateway in kfsystem.yaml.

Prevent creating a GSA policy when the GoogleProjectId field is not set.

Added Operator diagnostics to kf doctor.

Kf Operator to manage Kf installation.

Allow target command to take arg instead of flag.

Config Connector can manage the Kf Google Service Account (GSA).

Updated Tekton to 0.23.0.

Inline environment variable printing in kf env.

Removed internal routing dependency on internal-gateway.

Config Connector is now required.

Only check for timeout error for deprovisioning service instances.

Make targeting a non-existent Space an error.

Fixes manifest parsing bug.

Prevent panic in reconcilers when a Space is not found.

Allow long-running source uploads.

Kf supports Role-based access control in Spaces.

Kf supports source code upload without workstation write access to Artifact Registry.

Kf doctor supports running per-object commands.

Updated the ASM version to 1.9.1.

Ensure log tailing prints the correct number of lines.

Run binding before push so VCAP_SERVICES is correct on the first deploy.

Ensure a Space cannot be deleted if a ServiceInstance was not deleted.

Ensure a ServiceBroker cannot be deleted if there is still an active ServiceInstance depending on it.

Resolved issue with health-check-type=process.

Added support for Node Selector.

Added --as and --as-group global flags to support impersonation.

Make routes available in VCAP_APPLICATION .

Updated the Tekton version to 0.19.0.

Added feature flags enable_dockerfile_builds , enable_custom_buildpacks and enable_custom_stacks .

Made hostname as an optional field when creating a route.

Added support for Task.

Changed the kf CLI to use kubectl kubeconfig loading logic.

Added health-check-http-endpoint flag to kf push command.

Added support for Nodes to kf doctor .

Kf now shows a tip to run kf bindings after binding a service.

Added autoscaling capabilities to Kf.

kf doctor now checks for correct Istio injection labels on spaces managed by Kf.

Added support for the kf-domains subcommand so developers can look up which domains they can use in a Space. Operators can continue to use kf configure-space to modify domains.

Updated the Istio version to 1.7 and fixes the injection label.