Table of Contents generated with DocToc
Download v1.8.7 in the v1.8.7 release page.
None.
- The base image
alpine
now has been promoted fromalpine:3.20.1
toalpine:3.20.2
. (#5270)
Download v1.8.6 in the v1.8.6 release page.
karmada-controller-manager
: fix the issue of residual work in the MultiClusterService feature. (#5213, @XiShanYongYe-Chang)
None.
Download v1.8.5 in the v1.8.5 release page.
karmada-controller-manager
: Fixed the issue that the default resource interpreter doesn't accurately interpret the numbers of replicas. (#5106, @whitewindmills)
- The base image
alpine
now has been promoted fromalpine:3.20.0
toalpine:3.20.1
. (#5088)
Download v1.8.4 in the v1.8.4 release page.
None.
- The base image
alpine
now has been promoted fromalpine:3.19.1
toalpine:3.20.0
.
Download v1.8.3 in the v1.8.3 release page.
karmada-controller-manager
: Fixed incorrect annotation markup when policy preemption occurs. (#4770, @XiShanYongYe-Chang)karmada-controller-manager
: Fix the problem that labels cannot be deleted via Karmada propagation. (#4791, @whitewindmills)karmada-controller-manager
: Fix the problem that work.karmada.io/permanent-id constantly changes with every update. (#4820, @XiShanYongYe-Chang)
None.
Download v1.8.2 in the v1.8.2 release page.
karmada-search
: Add the logic of checking whether the resource API to be retrieved is installed in the cluster. (#4588, @XiShanYongYe-Chang)karmada-search
: support accept content typeas=Table
in the proxy global resource function. (#4594, @jwcesign)
- The base image
alpine
now has been promoted fromalpine:3.18.5
toalpine:3.19.1
. (#4599, @Fish-pro)
Download v1.8.1 in the v1.8.1 release page.
karmada-controller-manager
: fix incorrectforType
incluster-resource-binding-status controller
fromResouceBinding
toClusterResourceBinding
. (#4365, @lxtywypc)karmadactl
: Fixed return err in case ofsecret.spec.caBundle
is nil. (#4391, @XiShanYongYe-Chang)
- Karmada is now built with Go1.20.12. (#4397, @RainbowMango)
Download v1.8.0 in the v1.8.0 release page.
With the newly designed MultiClusterService
API, users can easily share services across clusters and the services can be consumed directly with the domain name.
See Service discovery with native Kubernetes naming and resolution for more details.
(Feature contributors: @bivas, @XiShanYongYe-Chang, @jwcesign, @Rains6)
With the optimized static weight replica assignment algorithm, the replicas can be evenly scheduled to desired clusters, this guarantees that resource consumption in large-scale scenarios is more balanced.
See Divide replicas by static weight evenly for more details.
(Feature contributors: @chaosi-zju)
This release significantly enhanced the project's security including the removal of insecure component access methods(see #4024), ensuring all components deployed by installation tools have secure configurations by default(see #4191), and so on.
(Feature contributor: @chaosi-zju, @yanfeng1992, @zhzhuang-zju)
Any resources deployed through Karmada can be protected to avoid accidental deletion. The resource template deployed in Karmada may exist widely across multiple clusters, accidental deletion of resource template can lead to serious accidents, such as accidental deletion of namespaces may cause all resources in the namespace to be cascaded removed. To eliminate this concern, this feature provides protection policies for any resource type, protected resources are not allowed to be removed.
See Resource Deletion Protection for more details.
(Feature contributor: @Vacant2333, @XiShanYongYe-Chang)
MultiClusterService
: IntroducedServiceProvisionClusters
andServiceConsumptionClusters
which will be used to specify service source and consumption place. (#4290, @jwcesign)
- The following label on
resource template
now has been deprecated and will be removed from the following releases:propagationpolicy.karmada.io/name
replaced bypropagationpolicy.karmada.io/permanent-id
propagationpolicy.karmada.io/namespace
replaced bypropagationpolicy.karmada.io/permanent-id
clusterpropagationpolicy.karmada.io/name
replaced byclusterpropagationpolicy.karmada.io/permanent-id
propagationpolicy.karmada.io/uid
replaced bypropagationpolicy.karmada.io/permanent-id
clusterpropagationpolicy.karmada.io/uid
replaced byclusterpropagationpolicy.karmada.io/permanent-id
These labels were used to refer toPropagationPolicy
orClusterPropagationPolicy
, now they are replaced by the permanent label.
karmada-aggregated-apiserver
: Fixed the issue that can not proxyexec
request to a proxy issue. (#4020, @XiShanYongYe-Chang)karmada-aggregated-apiserver
: Fix exec failure due to incorrupt TLS config. (#4206, @jwcesign)karmada-operator
: Fixed the issue that karmada-metrics-adapter was not removed after deleting a Karmada instance. (#4056, @wawa0210)karmada-operator
: Fixed can not load Karmada v1.7.0 crds issue. (#4130, @liangyuanpeng)karmada-operator
: Fix karmada instance aggregated service externalName and namespace error. (#4210, @wawa0210)karmada-operator
: Fixed an issue that components can not be deployed due to invalid manifests. (#4318, @calvin0327)karmada-operator
: resolve resource version conflict when updating service. (#4320, @calvin0327)karmada-webhook
: Fixed to validate spec.types of MultiClusterService API. (#4096, @lonelyCZ)karmada-controller-manager
: Fix panic when FederatedHPA's SelectPolicy is nil and FederatedHPA webhook is disabled. (#4103, @jwcesign)karmada-controller-manager
: Fixed a panic issue due toindex out of range
in resource model functionality. (#4145, @halfrost)karmada-controller-manager
: Fixed inconsistent generation issue betweenmetadata.generation
andstatus.observedGeneration
. (#4138, @jwcesign)karmada-controller-manager
: only updateaggregated status
andconditions
fields duringbinding-status controller
updating status of binding. (#4226, @lxtywypc)karmada-controller-manager
: Fix the bug that losing the chance to unclaim resource template in case of PropagationPolicy/ClusterPropagationPolicy is removed. (#4245, @whitewindmills)karmada-search
: Fixed can not access ResourceRegistry issue due to misconfigured singular name. (#4144, @zhzhuang-zju)karmada-search
: Fixed the ResourceRegistry is incompatible in the upgrade scenario. (#4171, @zhzhuang-zju)karmada-search
: Fix lock race affects watch RestChan not close, causing client watch api to hang. (#4212, @xigang)
- Introduced
trivy
for image security scanning. (#4123, @zhzhuang-zju) - bump golang.org/x/net to v0.17.0 to address CVE(CVE-2023-39325, CVE-2023-44487) concerns. (#4167, @RainbowMango)
- Bump golang.org/grpc to v1.56.3 to address security concerns(GHSA-m425-mq94-257g). (#4196, @zhzhuang-zju)
karmadactl
: Thekarmada-apiserver
andkarmada-aggregated-apiserver
installed by theinit
command will take--tls-min-version=VersionTLS13
by default. (#4181, @yanfeng1992)karmadactl
: Thekarmada-search
andkarmada-metrics-adapter
installed by theaddon
command will take--tls-min-version=VersionTLS13
by default. (#4193, @yanfeng1992)- The
karmada-apiserver
,karmada-aggregated-apiserver
,karmada-search
, andkarmada-metrics-adapter
installed bykarmada-operator
andchart
will take--tls-min-version=VersionTLS13
by default. (#4198, @zhzhuang-zju) - Fixed CVE-2016-2183 by setting Golang's secure cipher suites to ETCD's cipher suites. (#4253, @zhzhuang-zju)
karmadactl
: Theinit
now can setup production a production-ready HA control plane by using an external HA ETCD cluster. (#3898, @tedli)karmadactl
: Introduced--cluster-zones
flag toregister
command to declare cluster zones during registering a cluster. (#4069, @MingZhang-YBPS)karmadactl
: Introduced--proxy-server-address
flag toregister
command to declare proxy server during registering a cluster. (#4076, @yanfeng1992)karmadactl
: Introduced--cluster
flag to thetop
command to specify the cluster name. (#4223, @zhzhuang-zju)karmadactl
: Introduced--dependencies
flag topromote
command, which will be used to promote dependencies. (#4135, @zhy76)karmada-controller-manager
: Now thecurrentReplicas
anddesiredReplicas
in the status of HPA will be aggregated to the resource template by default. (#4064, @chaosi-zju)karmada-controller-manager
: IntroducedhpaReplicasSyncer
controller which syncs workload's replicas from the member cluster to the control plane. (#4072, @lxtywypc)karmada-controller-manager
: Now the replicas of deployments can be automatically retained if it is scaling with an HPA. (#4078, @chaosi-zju)karmada-controller-manager
: Fix panic caused by concurrent global variable reads and writes. (#4176, @jwcesign)karmada-controller-manager
: Ignoredconfigmap/extension-apiserver-authentication
from propagation. (#4237, @zhzhuang-zju)karmada-controller-manager
: Pruned job labelsbatch.kubernetes.io/controller-uid
andbatch.kubernetes.io/job-name
which were introduced by Kubernetes 1.27. (#4160, @liangyuanpeng)karmada-controller-manager
: Introduced permanent ID to PropagationPolicy, ClusterPropagationPolicy, ResourceBinding, ClusterResourceBinding, and Work. (#4199, @jwcesign)karmada-controller-manager
: Resource models now support any arbitrary resource type and are no longer limited tocpu
,memory
,storage
, andephemeral-storage
. (#4307, @chaosi-zju)karmada-controller-manager
: Introducedmulticlusterservice
controller to sync MultiClusterService. (#4323, @Rains6)karmada-agent
: Introduced--cluster-zones
flag to declare cluster zones during registering a cluster. (#4069, @MingZhang-YBPS)karmada-scheduler
: Introduced rate limiter options including:--rate-limiter-base-delay
,--rate-limiter-max-delay
,--rate-limiter-qps
, and--rate-limiter-bucket-size
, the default value not changed compared to the previous version. (#4105, @yanfeng1992)karmada-scheduler
: Fixed an issue that the scheduler does not ignore RB/CRB by scheduler name during enqueuing. (#4139, @yanfeng1992)karmada-scheduler
: Introduced leaderElection options including:--leader-elect-lease-duration
,--leader-elect-renew-deadline
,--leader-elect-retry-period
, the default value not changed compared to the previous version. (#4158, @yanfeng1992)karmada-scheduler
: The replicas now can be evenly distributed across clusters in case of scheduling replicas based on static weight. (#4219, @chaosi-zju)karmada-aggregated-apiserver
: Add ca to check the validation of the member clusters' server certificate. (#4183, @jwcesign)karmada-aggregated-apiserver
: Supports cross-cluster unified query. (#4254, @chaunceyjiang)karmada-operator
: Support config required resources for each Karmada component. (#4059, @wawa0210)karmada-operator
: Attached kubeconfig file into status(.status.secretRef) after installation. (#3789, @calvin0327)karmada-operator
: The default installation version of Karmada now is v1.7. (#4127, @MolisXYliu)karmada-operator
: The default installed version of Karmada now depends on the operator version. (#4133, @chaosi-zju)karmada operator
: Now able to install karmada on a remote cluster. (#3926, @calvin0327)
- Bump Golang version to Go1.20.10. (#4169, @vikash485)
- Karmada(v1.8) is now built with Go1.20.11. (#4334, @CoderTH)
- Kubernetes dependencies now has been bumped to v1.27.8. (#4328, @RainbowMango)
helm-chart
: Fixed the issue that can not installkarmada-search
. (#4034, @chaosi-zju)Helm chart
: Added helm index for 1.7 release. (#4051, @a7i)Helm chart
: Added helm index for 1.7.1 release. (#4178, @a7i)- Fixed render issue of podDisruptionBudget in controller-manager. (#4188, @a7i)
- Now able to install the component
karmada-metrics-adapter
with the chart. (#4303, @jwcesign)
karmada-controller-manager
: Skip emitting theApplyPolicyFailed
event for the resource template which does not match any PropagationPolicy. (#4052, @wu0407)
Thank you to everyone who contributed to this release!
Users whose commits are in this release (alphabetically by username)
- @22RC
- @a7i
- @calvin0327
- @chaosi-zju
- @chaunceyjiang
- @CoderTH
- @codrAlxx
- @huangyutongs
- @halfrost
- @JadeFlute0127
- @jwcesign
- @liangyuanpeng
- @lonelyCZ
- @lxtywypc
- @MingZhang-YBPS
- @MolisXYliu
- @RainbowMango
- @Rains6
- @tedli
- @Vacant2333
- @vikash485
- @watermeionG
- @wawa0210
- @whitewindmills
- @wrhight
- @wu0407
- @xigang
- @XiShanYongYe-Chang
- @Yan-Daojiang
- @yanfeng1992
- @zhy76
- @zhzhuang-zju