Releases: vmware-tanzu/tanzu-framework
Tanzu Framework v0.29.0
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Miscellaneous
-
Add VolumeSnapshot related RBACs to provider service account for TKC/GC (#4491, @akankshapanse)
-
Addon finalizer is added to workload clusters to control resource deletion order (#4160, @adduarte)
-
Addons manager self managed tls certificates have a life expectency of one week, and are rotated every six days. (#4317, @adduarte)
-
Allow for simpler ClusterClass JSON patches accessing
kubernetesSpec.{etcd,pause,coredns,kube-vip}.imageRepository fields in TKR_DATA values.
They can now assume imageRepository fields in TKR_DATA values are non-empty and default to
kubernetesSpec.imageRepository (if they are empty in the TKR). (#4298, @imikushin) -
Annotating a clusterbootstrap CR with tkg.tanzu.vmware.com/unmanaged-cni: "", will allow for the clusterbootstrap.spec.cni = nil (#4264, @adduarte)
-
Azure: Disable use of AAD Pod Identity for resolution of Azure credentials
- Azure: Restart pods during private cluster initialisation to work around lack of hairpin LB support for control planes
- Azure: Disable extraneous route table entries (#4273, @randomvariable)
-
Bump kapp-controller to v0.41.6 (#4374, @shivaani0505)
-
CSI will be deployed in namespace configure in vpshere-csi package. addon-controller will not manage vsphere-csi namespace. (#4512, @nikhilbarge)
-
Changed the capabilities deployment update strategy to fallback-on-replace (#4333, @yharish991)
-
Cluster-api-ipam-provider-in-cluster respects the imagePath field from the TKG BOM (#4529, @christianang)
-
Default vsphere csi namespace is changed from kube-system to vmware-system-csi (#4430, @nikhilbarge)
-
Deprecate TKG discovery functions in Capabilities (#4434, @codegold79)
-
Enable BUILDKIT caching for Tanzu Framework build (#3972, @jayunit100)
-
Fetch vSphere credentials from vSphereClusterIdentity, instead of secret directly, in case of TMC based deployment of vSphere CSI. (#4388, @akankshapanse)
-
Fix TKG_CUSTOM_IMAGE_REPOSITORY_SKIP_TLS_VERIFY user env setting for the TKR Source Controller. (#4383, @imikushin)
-
Fix issue where vsphere-cpi cannot set the node address after kubelet is configured with a different address on IPv6 clusters. (#4323, @christianang)
-
Fixed apigroups for volumesnapshot psa RBAC (#4547, @deepakkinni)
-
Fixes bug in clusterbootstrap webhook which incorrectly failed create validation if not all api service endpoints are running at the time of creation. (#4459, @adduarte)
-
Management cluster now uses cert-manager v1.10.2. (#4490, @avi-08)
-
Omits PCI related params from tkg vsphere config (#4457, @srm09)
-
Removed default FeatureGate CR belonging to config.tanzu.vmware.com API group in featuregates package (#4456, @yharish991)
-
Support trusting self-signed image registry for user's application. Users can configure 3 registries with a legacy config file:
ADDITIONAL_IMAGE_REGISTRY_1: ""
ADDITIONAL_IMAGE_REGISTRY_1_SKIP_TLS_VERIFY: false
#! Base64 encoded self-signed CA certificate
ADDITIONAL_IMAGE_REGISTRY_1_CA_CERTIFICATE: ""ADDITIONAL_IMAGE_REGISTRY_2: ""
ADDITIONAL_IMAGE_REGISTRY_2_SKIP_TLS_VERIFY: false
ADDITIONAL_IMAGE_REGISTRY_2_CA_CERTIFICATE: ""ADDITIONAL_IMAGE_REGISTRY_3: ""
ADDITIONAL_IMAGE_REGISTRY_3_SKIP_TLS_VERIFY: false
ADDITIONAL_IMAGE_REGISTRY_3_CA_CERTIFICATE: ""If there are more than 3, user can add them to generated cluster manifest variable additionalImageRegistries before creating the cluster. (#4467, @DanielXiao)
-
Tanzu CLI will throw error if current AVI controller version is not supported by current TKG version (#4338, @XudongLiuHarold)
-
Tanzu Cli will throw out errors when cluster name length is longer than the max allowed in avi enabled workload cluster. (#4416, @XudongLiuHarold)
-
Tanzu-auth package support for PSA (#4451, @benjaminapetersen)
-
This pr will enable tkgm user to confgure csi net-permissions (#4412, @nikhilbarge)
-
Updated Docker dependency in Go modules to v20.10.24 to fix CVE-2023-28840, CVE-2023-28841 and CVE-2023-28842. (#4549, @rajathagasthya)
-
Users can configure system-wide proxy when vsphere as the cloud provider (#4345, @XudongLiuHarold)
-
When WORKER_NODE_NAMESERVERS or CONTROL_PLANE_NODE_NAMESERVERS variables are set, nameservers from DHCP are ignored. (#4023, @tylerschultz)
Dependencies
Added
Changed
- github.com/docker/docker: v20.10.17+incompatible → v20.10.24+incompatible
- github.com/vmware/govmomi: v0.27.1 → v0.30.2
- golang.org/x/exp: e8c3332 → 540bb73
- golang.org/x/mod: v0.7.0 → v0.8.0
- golang.org/x/net: v0.4.0 → v0.8.0
- golang.org/x/sys: v0.3.0 → v0.6.0
- golang.org/x/term: v0.3.0 → v0.6.0
- golang.org/x/text: v0.5.0 → v0.8.0
- golang.org/x/tools: v0.1.12 → v0.6.0
- sigs.k8s.io/cluster-api-provider-vsphere: v1.4.1 → v1.5.3
Removed
- github.com/davecgh/go-xdr: e6a2ba0
New Contributors
- @akankshapanse made their first contribution in #4388
- @mohitsharma-in made their first contribution in #4382
- @nikhilbarge made their first contribution in #4412
- @joshuatcasey made their first contribution in #4514
- @deepakkinni made their first contribution in #4547
- @cfryanr made their first contribution in #4566
Full Changelog: v0.28.0...v0.29.0
Tanzu Framework v0.28.1
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Miscellaneous
- Addon finalizer is added to workload clusters to control resource deletion order (#4160, @adduarte)
- Allow for simpler ClusterClass JSON patches accessing
kubernetesSpec.{etcd,pause,coredns,kube-vip}.imageRepository fields in TKR_DATA values.
They can now assume imageRepository fields in TKR_DATA values are non-empty and default to
kubernetesSpec.imageRepository (if they are empty in the TKR). (#4298, @imikushin) - Annotating a clusterbootstrap CR with tkg.tanzu.vmware.com/unmanaged-cni: "", will allow for the clusterbootstrap.spec.cni = nil (#4264, @adduarte)
- Azure: Disable use of AAD Pod Identity for resolution of Azure credentials
- Azure: Restart pods during private cluster initialisation to work around lack of hairpin LB support for control planes
- Azure: Disable extraneous route table entries (#4273, @randomvariable)
- Bump kapp-controller to v0.41.6 (#4447, @shivaani0505)
- Fetch vSphere credentials from vSphereClusterIdentity, instead of secret directly, in case of TMC based deployment of vSphere CSI. (#4388, @akankshapanse)
- Fix TKG_CUSTOM_IMAGE_REPOSITORY_SKIP_TLS_VERIFY user env setting for the TKR Source Controller. (#4383, @imikushin)
- Tanzu CLI will throw error if current AVI controller version is not supported by current TKG version (#4338, @XudongLiuHarold)
- Upgrade CAPZ version to v1.6.3 (#4422, @srm09)
- Users can configure system-wide proxy when vsphere as the cloud provider (#4345, @XudongLiuHarold)
Dependencies
Added
Changed
- github.com/vmware/govmomi: v0.27.1 → v0.30.2
- golang.org/x/exp: e8c3332 → 540bb73
- sigs.k8s.io/cluster-api-provider-vsphere: v1.4.1 → v1.5.3
Removed
- github.com/davecgh/go-xdr: e6a2ba0
New Contributors
- @harisund made their first contribution in #4297
- @alfredthenarwhal made their first contribution in #4376
Full Changelog: v0.28.0...v0.28.1
Tanzu Framework v0.28.0
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Bug or Regression
- Fixed
tanzu management-cluster create --dry-run
showing plugin sync log message (#4106, @anujc25) - Install CLIPlugin CRD through core-management-plugins package (#4253, @anujc25)
Refactor
- CI: Refactor github workflow 'Main' to improve the workflow execution time (#4000, @navidshaikh)
- Create a separate go module and a top level seperate directory
cli/core
for Core CLI (#3302, @prkalle) - Move
apis/config/ClientConfig
andapis/cli/Catalog
API tocli/runtime/apis
module (#3375, @anujc25) - Refactor cmd/cli/plugin/package to have its own go.{mod,sum} and contain openapischema package (#3316, @codegold79)
- Refactored capabilities to create discovery and controller go module and moved them to a top level directory (#3260, @yharish991)
- Refactored codegen plugin and made it a go module (#3284, @yharish991)
Feature
- Added discovery source for TMC.
Fixed fetching artifacts over http to handle large file reads by processing chunks at a time. (#2950, @pgandigesang) - Update the Tanzu Feature CLI plugin and FeatureGate client to work with the new (core.tanzu.vmware.com v1alpha2) Feature and FeatureGate APIs. This change is a breaking change as the Tanzu Feature CLI will no longer work with config.tanzu.vmware.com v1alpha1. Please upgrade Feature and FeatureGate resources to use the new v1alpha2 CRDs. (#3850, @codegold79)
Failing Test
- Skips cc based integration tests if the PR is from a fork (#3313, @saji-pivotal)
Other
- Add new
node-role.kubernetes.io/control-plane
toleration for featuregates-manager. (#3489, @codegold79)
Miscellaneous
-
- Deprecates Capabilities API in run API group
- Adds new Capabilities API in core API group (#3474, @yharish991)
-
Add BYOI subcommand to import public OS image. (#3376, @HanFa)
-
Add
CONTROL_PLANE_NODE_LABELS
param to support adding labels for control plane nodes upon cluster creation. (#3818, @ydp) -
Add check to not clone from template if custom cluster bootstrap annotation is present (#3784, @maralavi)
-
Add control plane node selector and toleration to kappcontroller (#3537, @maralavi)
-
Add control plane toleration to metrics-server (#3553, @ggpaue)
-
Add control-plane tolerations to legacy kapp-controller and addons-manager (#3590, @shyaamsn)
-
Add default tolerations for addons-manager package values.yaml (#3560, @ggpaue)
-
Add flags to enable Kubevip as loadbalancer provider (#3846, @lubronzhan)
-
Add kube-vip-cp-config controller to support kube-vip as load balancer provider (#3845, @lubronzhan)
-
Add proxy configuration to tkrSourceControllerPackage when creating MC (#3751, @lubronzhan)
-
Add support for Node IPAM for vSphere. IP Addresses can be claimed from an IPAM provider instead of DHCP. (#3899, @christianang)
-
Add support for applying extra manifests during management cluster bootstrap. A directory of manifests can be applied to the tkg-system namespace just before cluster creation using
--additional-tkg-system-manifests
(#3886, @HanFa) -
Add the
node-role.kubernetes.io/control-plane
toleration to capabilities manager. (#3499, @seemiller) -
Added capability to deploy a cluster-class based management cluster for CAPD (#3193, @saji-pivotal)
-
Added cert-manager v1.9.1 manifests to the providers. (#3648, @sathyanarays)
-
Added the following new commands in package plugin (kctrl enabled):
-package init
- Author packages
-package release
- Release packages created usinginit
-package repository release
- Release package repositories
-package repository kick
- Kick a package repository (#3747, @praveenrewar) -
AddonConfig controllers enqueue Cluster by ownerRef (#3510, @lubronzhan)
-
Adds VSphere Template Resolver: Cluster Webhook which updates the template and moid in osImageRef after resolving the templates available on the target cluster.
- If the template is moved (the template path is changed) and the cluster is already resolved, users will need to manually edit the cluster and remove the corresponding
osImageRef.template
in the TKR Data. (#2748, @ShashankGirish)
- If the template is moved (the template path is changed) and the cluster is already resolved, users will need to manually edit the cluster and remove the corresponding
-
Adds logic to ensure addons-manager and kapp-controller aws ingress rules are present after upgrade (#4086, @adduarte)
-
Admission Webhook test to verify CA cert gets propagated to validating/mutating webhooks matching addons label selector (#3469, @maralavi)
-
Apply ClusterBootstrap objects from bootstrap cluster to management cluster (#3960, @shyaamsn)
-
Builder plugin no longer combines stdout & stderr when reading the plugin descriptor. (#3909, @garethjevans)
-
Bump CPI to 1.24.0 (#3231, @lubronzhan)
-
Bump kapp-controller to 0.41.2 (#3480, @haoranleo)
-
Bump kctrl in package plugin (#4249, @kumaritanushree)
-
CLI Plugin - isolated-cluster
Commands -
- download-bundle - Download images/bundle into local disk as TAR.
- upload-bundle - Upload images to private repository
Complete Command :
tanzu isolated-cluster download-bundle --source-repo --tkg-version --source-ca-certificate/--source-insecure
tanzu isolated-cluster upload-bundle --source-directory --destination-repo --destination-ca-certificate/--destination-insecure (#4103, @Vandy-P)
-
CNI:none enabled on ClusterClass Windows nodes. (#3867, @knabben)
-
Change the proxy setting in CC template (#3812, @lubronzhan)
-
Cluster API Provider Azure updated to v1.6.1 (#4255, @randomvariable)
-
Clusterbootstrap validation is handled different when cluster is marked by deletion.
Deleted resources do not fail validation (#3932, @adduarte) -
Contexts and current context are now stored in new config-ng.yaml (#3954, @mpanchajanya)
-
Create new top-level modules: tkr, object-propagation and util. (#3475, @imikushin)
-
Currently, all CLI plugins are specific to the Kubernetes context. As more services like TMC add plugins, there is some room for confusion as there is name collision between the plugins. To make the user experience very explicit, we are adding the target to the command invocation.
With this change, the Tanzu CLI supports two targets (context types):
kubernetes
,mission-control
. This is currently backwards compatible, i.e., the plugins are still available at the root level. In addition to that, we have all the contextual plugins grouped under their corresponding targets.Note: This is currently behind a feature flag. To enable the flag please run tanzu config set features.global.context-target true and then run
tanzu init
once after the upgrade.Please find some examples below:
- TKG workload cluster plugin:
Without target grouping (a TKG management cluster is set as the current active server)
~ tanzu cluster
With target grouping
~ tanzu kubernetes cluster
- TMC workload cluster plugin:
Without target grouping (a TMC server is set as the current active server)
~ tanzu cluster
With target grouping
~ tanzu mission-control cluster
(#3126, @giri-varma) -
Decouple Feature-Flags from
cli/runtime
library and move to it to plugin specific code (#3665, @anujc25) -
Default value of
features.package.kctrl-package-command-tree
is now set to true. To use old package plugin, usetanzu config set features.package.kctrl-package-command-tree false
(#3207, @praveenrewar) -
Deprecated Feature and FeatureGate API in config.tanzu.vmware.com, added new APIs Feature and FeatureGate in core.tanzu.vmware.com and a controller to reconcile feature.core.tanzu.vmware.com resources (#3613, @yharish991)
-
Enable Single Node Cluster deployment via tanzu CLI (#3674, @Ankitasw)
-
Enable pod security admission per default for Clusters in audit and warn mode. Also introduces variables to the ClusterClasses to make it configurable for ClusterClass based clusters. (#3862, @chrischdi)
-
Ensures aws ingress rules for addons-manager are in place after cluster upgrade (#4052, @adduarte)
-
Fix
current
field getting set to tmc context in client config (config.yaml) (#4090, @anujc25) -
Fix clusterboostrap webhook to include additional packages in CB but not present in CBT (#3896, @shyaamsn)
-
Fix clusterbootstrap continuous reconciliation (#4193, @shyaamsn)
-
Fix condition where management cluster creation may time out if kapp CRDs take longer to get installed. (#3713, @randomvariable)
-
Fix copyright headers to set year to 2023 (#4225, @shyaamsn)
-
Fix kapp-controller upgrade issue (#3631, @haoranleo)
-
Fix kapp-controller upgrade issue on workload cluster when package-based-lcm feature flag is off (#3663, @haoranleo)
-
Fix kapp-controller upgrade issue when package-based-lcm feature flag is on (#3700, @haoranleo)
-
Fix lint error (#3326, @haoranleo)
-
Fix lint issues under addons/
Fix lint issues under cmd/cli/plugin/packages (#3516, @adduarte) -
Fix package plugin tests flakiness (#3323, @haoranleo)
-
Fix package plugin tests for classy clusters (#4192, @shyaamsn)
-
Fix the bug in gen-publish-images-totar.sh when generating the command to copy
capabilities-package
andtkg-storageclass-package
(#3796, @danniel1205) -
Fix upgrade logic for additional packages in ClusterBoostrap but not in CBT (#3911, @shyaamsn)
-
Fixed a problem when resolving TKRs, where there are TKRs whose versions are longer
than the matching TKR's version.For example, the provided version prefix is
v1.22.13+vmware.1-tkg.1
. There are two TKRs
with versions satisfying this prefix:v1.22.13+vmware.1-tkg.1
,v1.22.13+vmware.1-tkg.1-zshippable
.
Of the ...
Tanzu Framework v0.25.4
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Dependencies
Added
Nothing has changed.
Changed
- sigs.k8s.io/cluster-api-provider-azure: v1.4.0 → v1.4.5
Removed
Nothing has changed.
Full Changelog: v0.25.3...v0.25.4
Tanzu Framework v0.25.3
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Feature
- Added discovery source for TMC.
Fixed fetching artifacts over http to handle large file reads by processing chunks at a time. (#2950, @pgandigesang)
Bug or Regression
Miscellaneous
-
Bump kapp-controller to 0.38.5 for release-0.25 (#3540, @haoranleo)
-
Currently, all CLI plugins are specific to the Kubernetes context. As more services like TMC add plugins, there is some room for confusion as there is name collision between the plugins. To make the user experience very explicit, we are adding the target to the command invocation.
With this change, the Tanzu CLI supports two targets (context types):
kubernetes
,mission-control
. This is currently backwards compatible, i.e., the plugins are still available at the root level. In addition to that, we have all the contextual plugins grouped under their corresponding targets.Note: This is currently behind a feature flag. To enable the flag please run tanzu config set features.global.context-target true and then run
tanzu init
once after the upgrade.Please find some examples below:
- TKG workload cluster plugin:
Without target grouping (a TKG management cluster is set as the current active server)
~ tanzu cluster
With target grouping
~ tanzu kubernetes cluster
- TMC workload cluster plugin:
Without target grouping (a TMC server is set as the current active server)
~ tanzu cluster
With target grouping
~ tanzu mission-control cluster
(#3126, @giri-varma) -
Currently, the Tanzu CLI supports storing authentication information on a per server basis (management cluster). This information is created using the login plugin and can be managed using the config server command in the core CLI. As we add more server types like a workload cluster, Tanzu Mission Control endpoint, etc. we need a consistent mechanism to manage the configuration for every server or rather for every context.
Also, we need to move away from the notion of server to contexts. To clarify, a server is just a remote endpoint that can be accessed via the Tanzu CLI (or its plugins). A context would include a server endpoint and any related client side configurations like default output format, feature flags enabled, etc. This can also be extended in future to support a local environment without any server if needed.
Hence, added a context command as part of the core CLI like below to create and manage contexts in the CLI.
Note: This is currently behind a feature flag. To enable the flag please run tanzu config set features.global.context-target true
Migration notes:
- CLI users should start using the
context
command instead of thelogin
command. - Plugin developers should use the
config.GetContext
methods in the core CLI library instead ofconfig.GetServer
when looking for server information.
~ tanzu context -h
Configure and manage contexts for the Tanzu CLIUsage:
tanzu context [command]Aliases:
context, ctx, contextsAvailable Commands:
create Create a Tanzu CLI context
delete Delete a context from the config
get Display a context from the config
list List contexts
use Make a context as the current active context for its typeFlags:
-h, --help help for contextUse "tanzu context [command] --help" for more information about a command. (#2202, @giri-varma)
- CLI users should start using the
-
Fixes UI bug that inserted trailing comma in cluster labels and annotations (#3662, @miclettej)
-
Users can now configure the NSX ALB controller with the same ALB resource name across multiple clouds. (#3721, @XudongLiuHarold)
Dependencies
Added
Nothing has changed.
Changed
- github.com/stretchr/objx: v0.2.0 → v0.4.0
- github.com/stretchr/testify: v1.7.1 → v1.8.0
- gopkg.in/yaml.v3: 496545a → v3.0.1
- sigs.k8s.io/cluster-api-provider-vsphere: v1.3.1 → v1.3.5
Removed
Nothing has changed.
Full Changelog: v0.25.1...v0.25.3
Tanzu Framework v0.25.1
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Enhancement
Bug or Regression
- Fix incorrect log message when creating management-cluster (#3172, @anujc25)
- Fixed "tanzu cluster get" command to return Classy cluster details on TKGS (#3049, @prkalle)
- Fixed an issue with Cluster UpdatesAvailable condition calculation:
when new TKRs (and their OSImages) are added to the system, they are not being considered
for UpdatesAvailable calculation for already existing clusters, until the clusters are
modified (e.g. updating an annotation) or the TKR Status Controller is restarted. (#3127, @imikushin) - Remove spaces in TKG_NO_PROXY; '*' is not allowed in TKG_NO_PROXY; Remove wrong comma in the end of TKG_NO_PROXY in AWS cluster settings (#3044, @lubronzhan)
- Use TKR v1alpha3 API commands based on feature-gate on the supervisor-cluster (#3057, @anujc25)
- Validate cluster kubernetes version upgrade check as initial step (#3033, @prkalle)
Miscellaneous
- Display correct error message when passing Cluster resource without ClusterClass as part of cluster create (#3076, @anujc25)
- Explicitly set criSocket for KubeadmControlPlane and KubeadmConfigTemplate in AWS and Azure to avoid race conditions at the node level (#3119, @yastij)
- Fixed
osInfo
annotation in Cluster CR is not correct after upgrade workload cluster with another os (#3092, @anujc25) - Fixes PCI passthrough support decimal conversion (#3080, @srm09)
- Implement logr wrapper as part of existing logger and use with clusterctl (#3122, @anujc25)
- Package-based-lcm: Fix data race in cluster pause webhook (#3055, @vijaykatam)
- The issue has been fixed in core addon's packages monitoring logic in creating workload cluster flow when cluster bootstrap does not have the correct package name (#3113, @chandrareddyp)
tanzu cluster create classy-cluster.yaml
does monitors addons core packages (kapp, cni, csi, cpi) installation after cluster creation (#3082, @chandrareddyp)
Dependencies
Added
Nothing has changed.
Changed
Nothing has changed.
Removed
Nothing has changed.
Tanzu Framework v0.26.0
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Refactor
- Create separate directory
cli/runtime
for cli integration library and make it a go module. (#3201, @anujc25) - Create separate go modules for
apis/cli
andapis/config
(#3141, @anujc25)
Enhancement
Bug or Regression
- A plugin error code is propagated to the core CLI.
When a command fails the error messages are no longer duplicated. (#3103, @marckhouzam) - Acquire lock before reading tanzu config file to avoid races with config writes (#3177, @anujc25)
- Any component using prompt.go will have the question mark color changed from green to cyan. (#3170, @mpanchajanya)
- Fix incorrect log message when creating management-cluster (#3171, @anujc25)
- Fixed "tanzu cluster get" command to return Classy cluster details on TKGS (#3050, @prkalle)
- Fixed an issue with Cluster UpdatesAvailable condition calculation:
when new TKRs (and their OSImages) are added to the system, they are not being considered
for UpdatesAvailable calculation for already existing clusters, until the clusters are
modified (e.g. updating an annotation) or the TKR Status Controller is restarted. (#3127, @imikushin) - Improved error messages when the user attempts to add a discovery source of an invalid type. (#3134, @marckhouzam)
- Package-based-lcm: Fix the issue of tkg-clusterclass packageinstall reconcile failed after 15 minutes of management-cluster creation (#3026, @anujc25)
- Remove spaces in TKG_NO_PROXY; '*' is not allowed in TKG_NO_PROXY; Remove wrong comma in the end of TKG_NO_PROXY in AWS cluster settings (#3044, @lubronzhan)
- Use TKR v1alpha3 API commands based on feature-gate on the supervisor-cluster (#3041, @anujc25)
- Validate cluster kubernetes version upgrade check as initial step (#3033, @prkalle)
Miscellaneous
-
Add Calico CNI E2E tests to AWS_CC suite (#3031, @codegold79)
-
Add TKGS E2E tests to verify the workload cluster is created successfully with Calico CNI (#2955, @HL-EverGreen)
-
Add automation tests for upgrade aws workload cluster (#3117, @maralavi)
-
Add automation tests for upgrade aws workload cluster with custom clusterbootstrap (#3178, @maralavi)
-
Add cc cluster E2E tests with custom clusterbootstrap on AWS (#3106, @ggpaue)
-
Add cc guest cluster upgrade E2E tests with CNI Calico on AWS (#3153, @HL-EverGreen)
-
Adds cluster upgrade integration test for tkgs (#3018, @adduarte)
-
Adds metadata versioning to the addons-manager package bundle (#3185, @adduarte)
-
Adds upgrade integration test for custom CB clusters (#3084, @adduarte)
-
Automate create/delete CC Guest cluster for a package with inline config on supervisor using Custom ClusterBootstrap and verify if the data value secret is created successfully on both supervisor and Guest cluster (#3146, @shivaani0505)
-
Automate create/delete CC Guest cluster on supervisor using Custom ClusterBootstrap and verify if all addons are successfully deployed (#3073, @shivaani0505)
-
Automate testing to check if addons are successfully deployed when guest cluster got created (#3036, @ggpaue)
-
Currently, the Tanzu CLI supports storing authentication information on a per server basis (management cluster). This information is created using the login plugin and can be managed using the config server command in the core CLI. As we add more server types like a workload cluster, Tanzu Mission Control endpoint, etc. we need a consistent mechanism to manage the configuration for every server or rather for every context.
Also, we need to move away from the notion of server to contexts. To clarify, a server is just a remote endpoint that can be accessed via the Tanzu CLI (or its plugins). A context would include a server endpoint and any related client side configurations like default output format, feature flags enabled, etc. This can also be extended in future to support a local environment without any server if needed.
Hence, added a context command as part of the core CLI like below to create and manage contexts in the CLI.
Note: This is currently behind a feature flag. To enable the flag please run tanzu config set features.global.context-target true
Migration notes:
- CLI users should start using the
context
command instead of thelogin
command. - Plugin developers should use the
config.GetContext
methods in the core CLI library instead ofconfig.GetServer
when looking for server information.
~ tanzu context -h
Configure and manage contexts for the Tanzu CLIUsage:
tanzu context [command]Aliases:
context, ctx, contextsAvailable Commands:
create Create a Tanzu CLI context
delete Delete a context from the config
get Display a context from the config
list List contexts
use Make a context as the current active context for its typeFlags:
-h, --help help for contextUse "tanzu context [command] --help" for more information about a command. (#2202, @giri-varma)
- CLI users should start using the
-
Display correct error message when passing Cluster resource without ClusterClass as part of cluster create (#3053, @anujc25)
-
Explicitly set criSocket for KubeadmControlPlane and KubeadmConfigTemplate in AWS and Azure to avoid race conditions at the node level (#3120, @yastij)
-
Fix vendir out of sync for antreaconfig crd (#3168, @maralavi)
-
Fixed
osInfo
annotation in Cluster CR is not correct after upgrade workload cluster with another os (#3091, @anujc25) -
Implement logr wrapper as part of existing logger and use with clusterctl (#3101, @anujc25)
-
Package-based-lcm: Fix data race in cluster pause webhook (#3055, @vijaykatam)
-
Package-based-lcm: Fix data race in clusterbootstrap webhook that tries to concurrently access GVR (#3064, @vijaykatam)
-
Plumb legacy skipCNIBinaries config to CalicoConfig API (#2774, @12345lcr)
-
The issue has been fixed in core addon's packages monitoring logic in creating workload cluster flow when cluster bootstrap does not have the correct package name (#3113, @chandrareddyp)
-
Version annotation has been added in ClusterClass templates (#3063, @chandrareddyp)
-
tanzu cluster create classy-cluster.yaml
does monitors addons core packages (kapp, cni, csi, cpi) installation after cluster creation (#3019, @chandrareddyp)
Dependencies
Added
Nothing has changed.
Changed
- github.com/AlecAivazis/survey/v2: v2.1.1 → v2.3.5
- github.com/Netflix/go-expect: 0e00d91 → 73e0943
- github.com/briandowns/spinner: v1.18.0 → v1.19.0
- github.com/cpuguy83/go-md2man/v2: v2.0.1 → v2.0.2
- github.com/creack/pty: v1.1.11 → v1.1.17
- github.com/google/go-cmp: v0.5.7 → v0.5.8
- github.com/hinshun/vt10x: d55458d → 820417d
- github.com/mattn/go-isatty: v0.0.14 → v0.0.16
- github.com/spf13/cobra: v1.4.0 → v1.5.0
- github.com/stretchr/objx: v0.2.0 → v0.4.0
- github.com/stretchr/testify: v1.7.1 → v1.8.0
- github.com/tj/assert: 0180943 → v0.0.3
- golang.org/x/sys: a9b59b0 → fbc7d0a
- gopkg.in/yaml.v3: 496545a → v3.0.1
- k8s.io/klog/v2: v2.40.1 → v2.60.1
Removed
Nothing has changed.
New Contributors
- @abhijit-dev82 made their first contribution in #2970
Full Changelog: v0.25.0...v0.26.0
Tanzu Framework v0.25.0
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Enhancement
- Check users' AVI related inputs valid or not (#2823, @XudongLiuHarold)
- Package-based-lcm: Update cluster upgrade and available-upgrades commands to support clusterclass based clusters (#2745, @prkalle)
Bug or Regression
- Add support to list clusterclass based clusters on TKGs (#2886, @prkalle)
- Fixed TKC based cluster template generation issue when
TKG_CUSTOM_IMAGE_REPOSITORY
is set (#2954, @chandrareddyp) - Fixes to validate workload cluster existence and throws an error if exists (#2987, @chandrareddyp)
Miscellaneous
-
AWS classy clusters now populate os-name information from config file and for all possible machineDeployments. (#2839, @tenczar)
-
Add AWS EBS CSI Driver and AzureDisk CSI Driver (#2875, @jeffwubj)
-
Add ClusterClass based cluster creation validation check at runtime based on feature-flag and feature-gate (#2804, @anujc25)
-
Add Object propagation controller to the tkg-clusterclass package (#2749, @anujc25)
-
Add node role selector to kapp controller (#2864, @maralavi)
-
Addons-manager "--cluster-delete-timeout" is exposed to templates as tanzuAddonsManager.deployment.clusterDeleteTimeout (#2996, @adduarte)
-
Adds the following variables to support PCI Passthrough on TKG on vSphere without Supervisor:
VSPHERE_WORKER_PCI_DEVICES, if set will configure PCI passthrough on all worker machines
VSPHERE_WORKER_PCI_DEVICES must be specified in the format <vendor_id>:<device_id>
e.g. VSPHERE_WORKER_PCI_DEVICES: "10DE:1EB8"VSPHERE_CONTROL_PLANE_PCI_DEVICES, if set will configure PCI passthrough on all control plane machines
VSPHERE_CONTROL_PLANE_PCI_DEVICES must be specified in the format <vendor_id>:<device_id>
e.g. VSPHERE_CONTROL_PLANE_PCI_DEVICES: "10DE:1EB8"VSPHERE_IGNORE_PCI_DEVICES_ALLOW_LIST, if set to true will allow other PCI devices to be passed through
other than NVIDIA T4 (i.e. "10DE:1EB8")VSPHERE_CONTROL_PLANE_CUSTOM_VMX_KEYS, if set will set custom VMX keys on all control plane machines
Keys must be set in the form =,=WORKER_ROLLOUT_STRATEGY configures the MachineDeployment rollout strategy. If set to OnDelete,
on updates, the existing worker machines will be deleted first before the replacement worker machines are created. (#2948, @randomvariable) -
Bug fix for regression caused by #2926 (#2974, @vijaykatam)
-
Bump Cluster API and providers (CAPI/CAPD to 1.1.5, CAPV provider to 1.3.1, CAPZ provider to 1.4.0) (#2951, @ykakarap)
-
Bump cert-manager to v1.7.2. (#2490, @ridavid2002)
-
Bump kapp-controller to 0.38.3 (#2719, @ridavid2002)
-
Bump kapp-controller to version 0.38.4 (#2880, @shivaani0505)
-
CAPI controller deployed during management cluster creation will activate ClusterTopology featuregate only if needed (#2994, @vuil)
-
Change default AVI_CONTROLLER_VERSION to 20.1.3. (#2931, @HanFa)
-
Enable leader election for Addons Controller (#2787, @HL-EverGreen)
-
Featuregates controller reconciles on Namespace resource changes. (#2984, @sathyanarays)
-
Fix incorrect namespace used with IsClusterClassBased during management cluster upgrade (#2859, @12345lcr)
-
Fixed object-propagation-controller copying cluster owned
MachineTemplate
andKubeadmConfigTemplate
resources to other namespaces (#3007, @anujc25) -
Fixes bug where passing a ccluster input file to create dry-run would error expecting a legacy config file. (#2899, @tenczar)
-
Kapp-controller .sh file update (#2901, @shivaani0505)
-
Legacy-style Clusters (with
tanzuKubernetesRelease
label) now have UpdatesAvailable status condition. If the condition is True, its message looks similar to the one in "classy" Clusters: a space-separated TKR version list, looking like this:[v1.21.3+vmware.1-tkg.3 v1.22.0+vmware.1-tkg.1 v1.22.0+vmware.1-tkg.2]
(#2997, @imikushin) -
Modifying machinehealthchecks on classy clusters is not supported. The API will return an error when performing these operations on classy clusters. (#2693, @tenczar)
-
Package-based-lcm: change cluster pause webhook behavior to not pause for legacy TKGS tkr's (#2935, @maralavi)
-
Package-based-lcm: set maxUnavailable to 1 in addons-manager deployment (#2962, @ggpaue)
-
Package-based-lcm: Add AntreaNSXPodRoutingEnabled to CPI CRD for uTKG routablePod (#2929, @lubronzhan)
-
Package-based-lcm: Fix stuck addons finalizer on deleted clusters (#2848, @vijaykatam)
-
Package-based-lcm: Verify addons installation and deletion as part of tkg clusterclass tests (#2913, @shyaamsn)
-
Package-based-lcm: add common functions to test addon packages (#2743, @ggpaue)
-
Package-based-lcm: fix memory leaks in packageinstallstatus_controller (#3004, @vijaykatam)
-
Package-based-lcm: updates node pool API to properly support all node pool customization values, e.g vmClass, storageClass (#2900, @tenczar)
-
Package-based-lcm: updates node pool API to support clusterclass based clusters doc (#2445, @tenczar)
-
Package-based-lcm: when feature flag is set, legacy style cluster creation should still work (#2898, @vuil)
-
Remove feature-gate validation for TKGS cluster and make it based on vSphere version (#2986, @anujc25)
-
Show core and additional package install statuses in clusterbootstrap status message field. (#2781, @codegold79)
-
Show plugin version as installed version and status based on discovered plugin version (#2903, @anujc25)
-
Support air-gapped scenario for context-scoped plugins to allow registry override (#2937, @anujc25)
-
Support configuring multiple default trusted registry for plugin downloads during build-time using
TANZU_PLUGINS_ALLOWED_IMAGE_REPOSITORIES
(#2977, @anujc25) -
The TKR Resolver Cluster webhook now honors the custom
imageRepository
cluster variable:
if present, itshost
field value is written into the resolved TKR_DATA values'
kubernetesSpec imageRepository fields. (#3013, @imikushin) -
Updates scale API to support clusterclass based clusters. (#2999, @tenczar)
-
Use fslock to acquire lock when reading/writing the tanzu config file for update (#2882, @anujc25)
-
When cluster's CNI implementation is Antrea, AKO will use NodePortLocal routing mode by default (#2938, @XudongLiuHarold)
Dependencies
Added
- github.com/Masterminds/squirrel: v1.5.2
- github.com/gofrs/uuid: v4.2.0+incompatible
- github.com/jmoiron/sqlx: v1.3.4
- github.com/lann/builder: 47ae307
- github.com/lann/ps: 62de8c4
- github.com/lib/pq: v1.10.4
- github.com/rubenv/sql-migrate: 55d5740
- gopkg.in/gorp.v1: v1.7.2
- helm.sh/helm/v3: v3.8.1
- oras.land/oras-go: v1.1.0
Changed
- github.com/Azure/aad-pod-identity: v1.8.6 → v1.8.9
- github.com/Azure/azure-sdk-for-go: v58.1.0+incompatible → v66.0.0+incompatible
- github.com/Microsoft/hcsshim: v0.8.24 → v0.8.21
- github.com/containerd/containerd: v1.5.11 → v1.5.9
- github.com/containerd/ttrpc: v1.1.0 → v1.0.2
- github.com/coredns/corefile-migration: v1.0.14 → v1.0.17
- github.com/cyphar/filepath-securejoin: v0.2.2 → v0.2.3
- github.com/docker/cli: v20.10.10+incompatible → v20.10.11+incompatible
- github.com/docker/docker: v20.10.12+incompatible → v20.10.16+incompatible
- github.com/hashicorp/consul/api: v1.11.0 → v1.12.0
- github.com/onsi/ginkgo/v2: v2.0.0 → v2.1.3
- github.com/onsi/gomega: v1.18.1 → v1.19.0
- github.com/sagikazarmark/crypt: v0.3.0 → v0.4.0
- github.com/spf13/viper: v1.10.0 → v1.10.1
- github.com/vmware-tanzu/carvel-vendir: v0.24.0 → v0.26.0
- github.com/xeipuuv/gojsonschema: ac52e68 → v1.2.0
- google.golang.org/api: v0.62.0 → v0.63.0
- k8s.io/api: v0.23.4 → v0.23.5
- k8s.io/apiextensions-apiserver: v0.23.4 → v0.23.5
- k8s.io/apimachinery: v0.23.4 → v0.23.5
- k8s.io/apiserver: v0.23.4 → v0.23.5
- k8s.io/client-go: v0.23.4 → v0.23.5
- k8s.io/code-generator: v0.23.4 → v0.23.5
- k8s.io/component-base: v0.23.4 → v0.23.5
- sigs.k8s.io/apiserver-network-proxy/konnectivity-client: v0.0.27 → v0.0.30
- sigs.k8s.io/cluster-api-provider-azure: v1.2.1 → v1.4.0
- sigs.k8s.io/cluster-api-provider-vsphere: v1.1.0 → v1.3.1
- sigs.k8s.io/cluster-api/test: v1.1.2 → v1.1.5...
Tanzu Framework v0.23.0
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Bug or Regression
Other (Cleanup or Flake)
-
Fixed TKR Source Controller (pkg/v2/tkr/controller/tkr-source): it now installs
TKR package contents.Changed TKR Source Controller (pkg/v2/tkr/controller/tkr-source): only compatible TKR BOMs and TKR packages are being fetched now.
Improved util/patchset.PatchSet: it is now safe for concurrent use and more efficient. (#2524, @imikushin)
Miscellaneous
-
- new FormUtils.addDynamicControl method
- usage of FormArray for labels
- Usage of Reactive Form Validations from Rxweb
- tkg-labels shared component (#2515, @mpanchajanya)
- Adds scaffolding for the autoscaler package (#2393, @saji-pivotal)
- Bump cluster-api-azure to v1.2.1 (#2247, @saimanoj01)
- Bumping Antrea to v1.5.2 on ClusterResourceSet (#2353, @knabben)
- Creates scaffolding to convert storage class to a carvel package (#1984, @saji-pivotal)
- Fixed #2543 TKR resolver blocks cluster update even when cluster is marked for deletion:
TKR Resolver now allows the request to pass through for Clusters with deletionTimestamp set. (#2545, @imikushin) - Package plugin now provides extended functionality for better observability and debugging.
tanzu config set features.package.kctrl-package-command-tree true
can be used to enable this new behaviour.- package installed includes status command which can be used to tail status of a package install.
- package installed includes pause and kick commands which can be used to pause the reconciliation of a package install or force trigger the reconciliation of a package install.
- The tailing behaviour for a package install is also present in kick command and has been added to package installed create or update.
- Breaking changes when feature flag is activated:
--verbose
global flag is removed.--create-namespace
flag is removed from all the commands (package repository add
,package repository update
,package installed create
,package installed update
andpackage install
)--poll-interval
and-poll-timeout
are renamed to--wait-interval
and--wait-timeout
in all the commands (package repository add
,package repository update
,package repository delete
,package install
,package installed create
andpackage installed update
.--create
flag is removed inpackage repository update
.- REPOSITORY and TAG columns are replaced by
SOURCE
column which consist of source type (eg imgpkg), repo url and tag inpackage repository list
. - In
package available list
, SHORT-DESCRIPTION is not displayed by default (--wide flag can be used to display it). Also, LATEST-VERSION column is removed. package available get
provides an additional table which lists the available versions for that package.--generate-default-values-file
is renamed to--default-values-file-output
inpackage available get
which accepts a filename where the default values need to be saved. Also, all the values are commented by default.--package-name
flag is renamed to--package
inpackage installed create
andpackage install
.--install
flag is removed in thepackage installed update
. (#2522, @praveenrewar)
- Package-based-lcm: Add permissions required by VsphereCPIConfig and VSphereCSIConfig controller to create ClusterRole for CAPV to reconcile ProviderServiceAccount (#2575, @shyaamsn)
- Package-based-lcm: Add docs for updated addons-manager (#2412, @vijaykatam)
- Package-based-lcm: Fix cluster pause webhook to set the temporary pause annotation on cluster's TKR label change (#2514, @maralavi)
- Package-based-lcm: Fix constant reloading of certificates of controller-runtime by only changing certs on filesystem if contents have changed. (#2584, @adduarte)
- Package-based-lcm: Fix patching rbac rules of capv aggregated cluster role (#2595, @vijaykatam)
- Package-based-lcm: Fix reinvocation policy add logs to cluster pause webhook (#2642, @maralavi)
- Package-based-lcm: Fix vsphereCSIConfig controller bug where multiple vsphereclusters are resolved for 1 capicluster (#2553, @peterochodo)
- Package-based-lcm: Ignore clusters tied to TKR with legacy tkr label (#2624, @adduarte)
- Package-based-lcm: Sync csi feature states from supervisor cluster to workload cluster (#2493, @peterochodo)
- Package-bassed-lcm: Enhanced TKR Source Controller to consider ConfigMaps labeled with the new label
run.tanzu.vmware.com/additional-compatible-tkrs: ""
to mark TKRs with versions listed
in such ConfigMaps as compatible. Only ConfigMaps in the namespace configured in the tkr-source-controller--namespace
command line parameter are considered. (#2608, @imikushin) - Support building and publishing multiple versions of specific package as part of package repository (#2578, @anujc25)
- TBA (#2581, @vuil)
- The builder command has been restructured and can now be used programmatically. (#1792, @stmcginnis)
Dependencies
Added
- github.com/go-logr/stdr: v1.2.2
- github.com/vmware-tanzu/carvel-kapp-controller/cli: 3beaddc
- go.opentelemetry.io/otel/exporters/otlp/internal/retry: v1.4.0
Changed
- bazil.org/fuse: 65cc252 → 371fbbd
- github.com/Azure/aad-pod-identity: v1.8.5 → v1.8.6
- github.com/Azure/go-autorest/autorest/adal: v0.9.16 → v0.9.18
- github.com/Azure/go-autorest/autorest/azure/auth: v0.5.8 → v0.5.10
- github.com/Azure/go-autorest/autorest: v0.11.21 → v0.11.23
- github.com/armon/go-radix: v1.0.0 → 7fddfc3
- github.com/benbjohnson/clock: v1.1.0 → v1.3.0
- github.com/cenkalti/backoff/v4: v4.1.1 → v4.1.2
- github.com/cppforlife/cobrautil: bb86e69 → c28a9f2
- github.com/google/martian: 25dcb96 → v2.1.0+incompatible
- github.com/hashicorp/consul/sdk: v0.8.0 → v0.1.1
- github.com/hashicorp/go-multierror: v1.1.0 → v1.0.0
- github.com/hashicorp/mdns: v1.0.1 → v1.0.0
- github.com/hashicorp/memberlist: v0.2.2 → v0.1.3
- github.com/mattn/go-shellwords: v1.0.10 → v1.0.3
- github.com/miekg/dns: v1.1.26 → v1.0.14
- github.com/mitchellh/cli: v1.1.0 → v1.0.0
- github.com/posener/complete: v1.2.3 → v1.1.1
- github.com/rogpeppe/go-internal: v1.5.2 → v1.3.0
- github.com/smartystreets/assertions: v1.2.0 → b2de0cb
- github.com/smartystreets/goconvey: v1.7.2 → v1.6.4
- github.com/vmware-tanzu/carvel-kapp-controller: v0.25.0 → v0.35.0
- go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp: v0.25.0 → v0.27.0
- go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracegrpc: v1.0.1 → v1.4.0
- go.opentelemetry.io/otel/exporters/otlp/otlptrace: v1.0.1 → v1.4.0
- go.opentelemetry.io/otel/exporters/prometheus: v0.24.0 → v0.27.0
- go.opentelemetry.io/otel/internal/metric: v0.24.0 → v0.27.0
- go.opentelemetry.io/otel/metric: v0.24.0 → v0.27.0
- go.opentelemetry.io/otel/sdk/export/metric: v0.24.0 → v0.20.0
- go.opentelemetry.io/otel/sdk/metric: v0.24.0 → v0.27.0
- go.opentelemetry.io/otel/sdk: v1.0.1 → v1.4.0
- go.opentelemetry.io/otel/trace: v1.0.1 → v1.4.0
- go.opentelemetry.io/otel: v1.0.1 → v1.4.0
- go.opentelemetry.io/proto/otlp: v0.9.0 → v0.12.0
- golang.org/x/exp: e8c3332 → 6cc2880
- honnef.co/go/tools: v0.2.0 → v0.0.1-2020.1.4
- k8s.io/kube-aggregator: v0.19.2 → v0.22.4
- sigs.k8s.io/cluster-api-provider-azure: v1.0.1 → v1.2.1
- sigs.k8s.io/cluster-api/test: v1.1.1 → v1.1.2
Removed
- code.gitea.io/sdk/gitea: v0.12.0
- contrib.go.opencensus.io/exporter/aws: 2befc13
- contrib.go.opencensus.io/exporter/stackdriver: v0.12.1
- contrib.go.opencensus.io/integrations/ocsql: v0.1.4
- contrib.go.opencensus.io/resource: v0.1.1
- github.com/Azure/azure-amqp-common-go/v2: v2.1.0
- github.com/Azure/azure-pipeline-go: v0.2.2
- github.com/Azure/azure-service-bus-go: v0.9.1
- github.com/Azure/azure-storage-blob-go: [v0.8.0](https://github.com/Azure/azure-storage-blob-go/tree/v0....
Tanzu Framework v0.22.0
Tanzu CLI Installation Instructions
If you are installing Tanzu CLI using the artifacts published as part of this release, please follow the instructions.
Changes by Kind
Enhancement
Feature
- Package-based-lcm: Added support to configure deployment and webhook ports in tkr-service package (#2299, @prkalle)
Bug or Regression
- Deleting management cluster can run successfully even in high latency network env. (#2329, @XudongLiuHarold)
- Package-based-lcm: Fixed the indentation for Issuer resource in tkr-service package (#2319, @prkalle)
Miscellaneous
- Addons finalizers will not be added to clusters labeled as management clusters (#2425, @adduarte)
- Bump calico to 3.22.1 (#2229, @12345lcr)
- Cluster creation validates SERVICE_CIDR size does not exceed
kube-apiserver
's size restriction, and fails fast. IPv4 CIDRs mask must be/12
or smaller and IPv6 CIDRs must be/108
or smaller. (#2342, @tylerschultz) - Consider additional metadata as part of kubernetes build version during version comparison (#2289, @anujc25)
- Expose AVI_LABELS for creating workload cluster, for example
"{foo: 'bar'}"
, double quotations are required. (#2337, @HanFa) - Fix context-scoped plugins not getting installed after creating management-cluster through Kickstart UI (#2287, @anujc25)
- Fixed bug where ENABLE_CEIP_PARTICIPATION config variable was not being heeded (#2290, @saji-pivotal)
- Package-based-lcm:
Allow adding tkg.tanzu.vmware.com/add-missing-fields-from-tkr= in the annotation of ClusterBootstrap to enable the defaulting webhook logic which adds defaults to the missing spec fields and complete the partially filled ClusterBootstrapPackage RefName (#2403, @danniel1205) - Package-based-lcm:
- vsphere cpi config controller creates either paravirt or non-paravirt config
- ProviderServiceAccount for vsphere cpi and csi uses VsphereCluster in spec and ownerRef (#2198, @vijaykatam)
- Package-based-lcm: Adding tkr-status-controller to tkr-service package (#2322, @saji-pivotal)
- Package-based-lcm: Change addons-manager-v2 package name to addons-manager (#2315, @shyaamsn)
- Package-based-lcm: Create aggregate cluster role for CAPV (#2494, @vijaykatam)
- Package-based-lcm: Ensure addons-manager package has latest CRDs (#2507, @vijaykatam)
- Package-based-lcm: Fix VSphereCPIConfig data values type inconsistency and directly use endpoint IP. (#2196, @HanFa)
- Package-based-lcm: Fix
cliplugins
package deletion issue (#2400, @anujc25) - Package-based-lcm: Made healthz & webhook server port configurable in addons-manager package (#2354, @ggpaue)
- Package-based-lcm: Packaging tkr-source-controller (#2355, @prkalle)
- Package-based-lcm: Remove owner references from PackageInstall on workload clusters. (#2351, @adduarte)
- Package-based-lcm: Removes superfluous logging of machine-controller (#2331, @adduarte)
- Package-based-lcm: TKR Source Controller makes TKRs published as Carvel packages available on the management cluster. (#2173, @imikushin)
- Package-based-lcm: Update CalicoConfig
vethMTU
to string and removenamespace
(#2164, @12345lcr) - Package-based-lcm: Update SkipCNIBinaries param to CalicoConfig and fix CRD additionalPrinterColumns (#2416, @12345lcr)
- Package-based-lcm: addons-manager bug fixes (#2297, @vijaykatam)
- Package-based-lcm: postpone connecting to remote cluster until when needed (#2402, @maralavi)
- Package-based-lcm: set vsphere csi replica count based on controlplane node count (#2311, @peterochodo)
- Removed packages/{management,standalone} directories and move their contents to the packages directory. (#1906, @codegold79)
- Support deletion of IPv6 primary dualstack management clusters with automatically provisioned cleanup clusters. (#2348, @adobley)
- Update the KR plugin to implement the new TKR API v1alpha3 (#2242, @prkalle)
tanzu cluster
andtanzu management-cluster
commands and their sub-commands, do not print usage information in case the command fails for any reason, it just prints error information. To get usage information fortanzu cluster
ortanzu management-cluster
commands and their sub-commands, need to call the command or sub-commands with the flag--help
or-h
. (#2253, @chandrareddyp)
Dependencies
Added
Nothing has changed.
Changed
- github.com/gorilla/websocket: v1.4.2 → v1.5.0
- golang.org/x/sys: 22a9840 → a9b59b0
Removed
Nothing has changed.
Full Changelog: v0.21.0...v0.22.0