v0.3.0

Kamaji v0.3.0 can be installed using its v0.12.0 Helm Chart.

Remember to add the CLASTIX repository and update it:

helm repo add clastix https://clastix.github.io/charts
helm repo update

Images can be pulled from Docker hub:

docker pull clastix/kamaji:v0.3.0

⚠️ Due to several changes on the TenantControlPlane definition, ensure to update it manually if you've priorly installed Kamaji using Helm.

Enhancements

• Docs enhancement #298 @bsctl
• TCP deployment mangling (additional volumes, init containers, containers, and volume mounts) #223 @maruina @prometherion
• Support for TCP deployment Dynamic Resource Allocation #293 @prometherion
• Custom container registry #299 @prometherion

Hotfixes

• Wrong namespace in the Deployment guide #295 @gionn

v0.2.3

Container images are hosted on docker.io and can be pulled with the following command:

docker pull clastix/kamaji:v0.2.3

The current release has been tested over the last Chart release (0.11.5), remember to update your Helm chart as follows:

helm repo add clastix https://clastix.github.io/charts
helm repo update

Hotfixes

• CrashLoopBackOff when defining ingress in spec.controlPlane #282 @sn4psh0t @prometherion
• Missing Ingress loadbalancer status #283 @prometherion @sn4psh0t

Enhancements

• Support up to Kubernetes 1.27 #289 @bsctl @prometherion

v0.2.2

Container images are hosted on docker.io and can be pulled with the following command:

docker pull clastix/kamaji:v0.2.2

The current release has been tested over the last Chart release (0.11.4), remember to update your Helm chart as follows:

helm repo add clastix https://clastix.github.io/charts
helm repo update

Hotfixes

• Upon restoration of a TCP, Kamaji is not able to reconcile Deployment #279 @prometherion

v0.2.1

Container images are hosted on docker.io and can be pulled with the following command:

docker pull clastix/kamaji:v0.2.1

The current release has been tested over the last Chart release (0.11.3), remember to update your Helm chart as follows:

helm repo add clastix https://clastix.github.io/charts
helm repo update

Enhancements

• Code optimization for resource checksum #258 @prometherion
• Resource label consistency #269 @prometherion @bsctl
• Upgrading to new Kubernetes container registry #262 @ptx96
• Supporting Kubernetes v1.26.1 #274 @prometherion

Hotfix

• Support Velero restore #270 @ptx96 @prometherion

v0.2.0

Container images are hosted on docker.io and can be pulled with the following command:

docker pull clastix/kamaji:v0.2.0

The current release has been tested over the last Chart release (0.11.0), remember to update your Helm chart as follows:

helm repo add clastix https://clastix.github.io/charts
helm repo update

⚠️ Due to several changes on the TenantControlPlane definition, ensure to update it manually if you've priorly installed Kamaji.

Enhancements

• cert-manager support #166 (@prometherion)
• Preventing TCP downgrading via validating webhook #193 (@prometherion)
• Upgrade to latest Kubernetes release #196 #225 (@prometherion)
• Showing used datastore in kubectl get tcp output #183 (@bsctl @prometherion)
• DataStore migration support when using the same driver #164 (@prometherion)
• Removing rbac-proxy for metrics scraping #192 (@prometherion)
• Avoiding switch to a different DataStore driver #194 (@prometherion)
• Validating selected DataStore on TCP basis #206 (@prometherion)
• Freezing TCP during migration #208 (@prometherion)
• Tenant Control Plane addons reactive reconciliation #213 #218 (@prometherion)
• Updating GH Action base image #227 (@prometherion)
• Various reorgs #222 (@prometherion)
• Support installation on AWS EKS #187 (@maruina)
• Updating default etcd installation preventing segfault #202 (@bsctl @prometherion)
• Documenting FluxCD integration with Tenant Control Planes #121 (@curx @maxgio92)
• TCP Deploymentupdate strategy #242 (@bsctl @prometherion)
• Customizing kubelet preferred address types #245 (@fchiaretta @prometherion)
• First benchmark results #252 #249 (@prometherion)
• Documentation alignment #236 (@bsctl)
• ServiceMonitor support #251 (@prometherion)
• Performance enhancements #255 (@prometherion)

Hotixes

• Missing metadata upon container image release #190 (@prometherion)
• Failing reconciliation due tom missing secret #189 (@prometherion @maxgio92)
• Corruption of etcd prefix upon deletion of single TCP #201 (@bsctl @prometherion)
• DataStore certificate change truncate etcd prefix #204 (@prometherion)
• CA rotation with certs regeneration #240 (@prometherion)
• Handling etcd Helm chart basic auth values #254 (@bsctl)

Thanks

To all the contributors, issue reporters, or feature requesters, ad maiora! 🥂

v0.2.0-rc0

• c905e16 chore(docs/guides): fix syntax on flux helmrelease
• e08792a chore(docs/images): update flux diagram
• 248b508 docs(docs/content/guides/kamaji-gitops-flux.md): use third person
• 5cebb05 docs: add guide for managing tenant resources gitops way
• efbefba docs(api): aligning to latest changes
• bc19203 chore(gi): checking diff and docs alignement
• c8b8dcc test(e2e): testing different datastores and migration
• cf27212 chore: adding samples and automating deployment of datastores
• 4aa7792 chore(helm): using webhooks for secrets instead of finalizers
• a3c52e8 chore(kustomize): using webhooks for secrets instead of finalizers
• 7ed3c44 refactor(datastore): using webhooks for secrets instead of finalizers
• beebaf0 fix(storage): wrong variable while assigning finalizers
• b9cda29 fix(migrate): allowing leases updates during migration
• 7db8a64 fix(etcd): using stored username for cert common name
• c6abe03 fix(soot): typo on params for service name and namespace
• 723fa1a chore(helm): upgrading etcd to v3.5.6
• 7e0ec81 deps: upgrading etcd to v3.5.6
• 7353bb5 chore(gh): upgrading to go 1.19
• 96cedad test(e2e): upgrading to ginkgo v2
• 76b603d chore(helm): upgrade to 1.26
• 7cff6b5 chore(kustomize): upgrade to 1.26
• 6e6ea01 refactor(k8s): upgrade to 1.26
• aefdbc9 deps(k8s): upgrade to 1.26
• 074279b chore(go): upgrading to 1.19 required for k8s 1.26
• 09891f4 chore(gh): running workflows on ubuntu-22.04
• 18c6046 refactor: conforming finalizers management
• ceab662 feat(soot): using finalizer for clean-up
• d38098a fix(soot): ensure that manager is stopped upon tcp deletion
• 017a50b fix(soot): ensuring manager to restart upon tcp pod restart
• b07062b chore(kustomize): missing datastore finalizer rbac
• b880cff fix: missing datastore finalizer rbac
• 3f7fa08 refactor: removing unused scheme
• 8311f1f fix: ensure default datastore exists before starting manager
• 77fff03 chore(helm): support for runtime class
• abada61 chore(kustomize): support for runtime class
• 1eb1e0f feat: support for runtime class
• e83c347 refactor(soot): creating channel source during controller setup
• 3b90294 chore(helm): kubeadm phases are moved to soot manager
• c5d62f3 chore(kustomize): kubeadm phases are moved to soot manager
• 938341a refactor(log): uniforming log for soot controllers
• 3ea721c feat(kubeadm): moving phases to soot manager
• 5cbd085 chore(helm): addons no more need checksum
• e358fbe chore(kustomize): addons no more need checksum
• 9d55e77 refactor(api): no more need of checksum for addons
• 1e4640e feat(addons): implementation in the soot cluster
• 1b14922 refactor(kubeadm): preparing migration for addons to soot manager
• 11f8000 fix(konnectivity): typo in ca-cert cli flag
• e11b459 fix(konnectivity): reconciliation failed and in loop
• 9c8de78 docs: datastore migration support
• 4c51eaf feat(konnectivity): reconciliation performed by soot manager
• 1a80fc5 fix(api): wrong konnectivity defaults
• 02052d5 fix(helm): wrong konnectivity defaults
• 7e47e33 fix(kustomize): wrong konnectivity defaults
• 28c47d9 refactor: moving migrate webhook handling from tcp to soot manager
• 1ec257a feat: introducing soot controllers manager
• 68006b1 fix(datastore): coalesce for storage configuration
• cd109dc fix: using slash prefix for etcd datastore
• 1138eb1 fix: using the status storage schema for the etcd prefix
• f4f9140 feat(migrate): enhancing job metadata
• 5e78b63 feat(migrate): making timeout configurable
• e25f95d feat(migrate): making image configurable
• 7f49fc6 refactor(konnectivity): removing default logging options
• 8b96838 fix: support for arguments without a value
• cb5e356 docs: support for konnectivity extra args
• 0d6246c chore(helm): support for konnectivity extra args
• d8760fd chore(kustomize): support for konnectivity extra args
• c00df62 feat(konnectivity)!: support for extra args
• 653a393 chore(helm): decoupling agent and server struct
• 6775b2a chore(kustomize): decoupling agent and server struct
• 5241fa6 refactor(konnectivity)!: decoupling agent and server structs
• 723fef5 feat(migrate): injecting webhook into tcp
• 8d1d859 refactor: moving datastore migrate resource to its module
• c96f589 fix(helm): installing datastore upon completion
• 2d1daa8 feat(datastore): validation webhook
• fe94829 chore(helm): wrong crd validation markers
• 79942dd chore(kustomize): wrong crd validation markers
• 4491959 fix(kubebuilder): wrong crd validation markers
• 2336d40 refactor: using custom validator and custom defaulter
• 79c59e5 feat: validation webhook to prevent DataStore migration to a different driver
• 95d0983 chore(dockerfile): optimizing build
• 7e276e5 chore(helm): support to datastore migration w/ the same driver
• b2e6460 fix(helm): switching over webhook server service
• 3850ad9 chore(kustomize): support to datastore migration w/ the same driver
• 9e89937 feat: support to datastore migration w/ the same driver
• a260a92 fix(psql): checking db and table ownership
• cc4864c feat: datastore migration drivers
• ece1a4e fix: avoiding inconsistency upon tcp retrieval and status update
• eb2440a refactor: abstracting datastore configuration retrieval
• 0c41570 fix(datastore): not deleting database content upon certificates change
• 7a6b0a8 fix(datastore): ensuring to update status upon any change
• a31fbdc chore(makefile): allowing creation of multiple datastore instances
• 4ff0cdf docs: configuration for the manager command
• ae573b1 chore(kustomize): removing rbac proxy and support for manager command
• e81b322 chore(helm): removing rbac proxy and support for manager command
• 4298bdd chore(dockerfile): manager command
• 15d0d57 feat: refactoring for commands
• c17a31e fix: avoiding collision of datastore schemes
• f0df1cf fix: removing tcp data using prefix, and not range
• 1bcff90 chore(kustomize): show datastore for each tcp
• 6c817a9 chore(helm): show datastore for each tcp
• 5b9311f feat: show datastore for each tcp
• 0d607df refactor: adding finalizer upon datastore setu
• 11502bf refactor: retry on conflict for the status update
• ff1c9fc chore(samples): updating to latest kubeadm supported version
• adc4b7d chore(test): updating to latest kindest/node version
• a96133f deps: upgrade to k8s 1.25.4
• 81fb429 test(e2e): validating tcp kubernetes version
• 190acc9 feat: tcp version validation upon create and update
• b0a059d docs: cert-manager dependency
• bcc7d0e chore(makefile): installing cert-manager for e2e
• 9dc0a9a chore(makefile): crds diverged between kustomize and helm
• d312738 chore(helm): support for cert-manager and webhooks
• 30bc8cc feat!: support for cert-manager and webhooks
• 55d7f09 chore(kustomize): support for cert-manager and webhooks
• 2c892d7 fix(ci): missing metadata upon container images release
• 43f1a6b chore(makefile): installing required dependencies
• 78ef34c fix(docs): aligning to latest changes for the chart documentation
• 16d8b2d fix(helm): support installation on EKS
• 68764be chore(helm): support installation using --wait option

v0.1.1

Container images are hosted on docker.io and can be pulled with the following command:

docker pull clastix/kamaji:v0.1.1

The current release has been tested over the last Chart release (0.10.0), remember to update your Helm chart as follows:

helm repo add clastix https://clastix.github.io/charts
helm repo update

⚠️ Due to an update on the TenantControlPlane definition, ensure to update it manually if you've prior installed Kamaji.

Features

• Specify scheduling constraints for TenantControlPlane pods #179

Enhancements

• Kamaji listing on Rancher application catalog #170
• CNCF compliance support #169
• Docs link to nodes joining process #175
• Link to deploy guide #176

v0.1.0

Container images are hosted on docker.io and can be pulled with the following command:

docker pull clastix/kamaji:v0.1.0

Enhancements

• Reflecting current version of the deployed Tenant Control Plane version rather then the specified one in the spec [#5]
• Moving to Docker Hub as main container registry [#8] and GH Actions workflow [#16]
• Support for Go 1.18 [#12]
• Dynamic management of Addons (kube-proxy, core-dns) [#7, #107]
• Helm Chart automatic release [#50]
• Using default values for the network profile section of Tenant Control Plane [#58]
• Remove API Machinery NewYAMLSerializer() deprecated function [#60
• Report Kamaji version and info at startup [#62]
• Adding konnectivity as Addons [#25, #93, #100, #140]
• Missing or wrong header in source files [#71]
• Utils for tenant kubernetes client [#72]
• Making optional spec.networkProfile.domain and renaming it to certSANs [#75]
• Missing handling of conflicts during reconciliation [#54]
• Set image pull policy to Always [#82]
• Provide logs for deletion of Tenant Control Planes [#83]
• Make container resources allocation configurable in Tenant Control Plane pods [#99]
• Support for additional extra arguments in TCP components [#103]
• Support for etcd storage class and size in Helm chart [#112, #114]
• Introduce support for MySQL and PostgreSQL as datastore [#22]
• DataStore API [#67, #133]
• Upgrade dependencies to Kuberentes 1.25 [#136]
• Dynamic kube-proxy addon version [#137]
• DataStore selection at the Tenant Control Plane level [#138]
• Enable topology spread constraints on tcp deployment [#142]
• Add the scale subresource in TenantControlPlane custom resource definition [#144]

Bug fix

• Tenant Control Plane kubelet configuration is not updated upon cgroup driver change [#79]
• Reconciliation of Kubeconfig upon TCP address change [#90]
• kube-proxy kubeconfig is missing address when dynamically allocation [#97]
• Update link to Capsule in README [#120]
• Wrong ServiceAccountIssuer string in kube-apiserver [#149]
• Documentation alignements [#162]

What's next

Check the just opened milestone and feel free to open a feature request!

Thanks

As the first minor release of Kamaji we're really emotional and we would like to thank all the contributors and early adopters: @bsctl, @ptx96 , @stek29, @davideimola, @alegrey91, @maxgio92, @mendrugory, @prometherion.

v0.1.0-rc0

• reorg: upgrade phase as first resource by @prometherion in #6
• chore: using docker hub as default container registry by @prometherion in #9
• Support for Go 1.18 by @prometherion in #12
• reorg: scaffolding for e2e test suite by @prometherion in #15
• docs: pointing to k8s slack channel by @prometherion in #13
• Addons by @mendrugory in #7
• chore(helm): addons by @mendrugory in #24
• docs(kind): fixed URL for kamaji develop deploy by @ptx96 in #33
• CI - Build and push amd/arm OCI images by @ptx96 in #17
• fix(helm): missing crd update upon addons feature by @prometherion in #36
• test(e2e): add test suite for e2e by @alegrey91 in #34
• Update documentation with latest changes by @bsctl in #44
• reorg: marking loadbalancer errors as debug by @prometherion in #39
• test(e2e): worker node join through kubeadm by @prometherion in #49
• Update the documentation by @maxgio92 in #47
• chore(ci): helm release pipeline by @davideimola in #50
• fix(docs): wrong links by @bsctl in #53
• Using default values for the network profile section of Tenant Control Plane by @prometherion in #58
• refactor(kubeadm): using suggested serializer with options by @prometherion in #60
• feat: providing binary version and details by @prometherion in #62
• Refactor and konnectivity by @mendrugory in #25
• kine by @mendrugory in #66
• style: ensuring headers are correct by @prometherion in #71
• refactor: abstracting tenant control plane client generation by @prometherion in #72
• fix: konnectivity reconciliation loop by @prometherion in #74
• Support for additional certificate SANs by @prometherion in #75
• refactor: retrying mutate function in case of conflict error by @prometherion in #77
• Always pulling images by @prometherion in #82
• feat: support for announced clusterip address by @prometherion in #81
• Ensuring kubeadm phases and addons reconciliation by @prometherion in #79
• feat: logging resource deletion by @prometherion in #83
• Refactoring getting-started docs and automations by @maxgio92 in #68
• Embed etcd installation using Helm by @prometherion in #84
• fix(internal/resources): close etcd client after etcdsetup reconcile by @maxgio92 in #88
• Ensuring kubeconfig upon TCP change by @prometherion in #90
• fix(internal/resources/konnectivity): get lb svc endpoint from cp status by @maxgio92 in #93
• Support for container resources for each component by @prometherion in #99
• Using TCP status as address source of truth by @prometherion in #97
• Collapsing TCP and Konnectivity Service in a single resource by @prometherion in #100
• Additional extra arguments for TCP components by @prometherion in #103
• CLI flag to specify Kine container image by @prometherion in #104

v0.0.1

Container images are available in the following container registries.

docker pull clastix/kamaji:v0.0.1
docker pull quay.io/clastix/kamaji:v0.0.1

Features

• Support for dynamic address allocation on native Load Balancer
• Zero Downtime Tenant Control Plane upgrade
• Helm Chart to install Kamaji on any Kubernetes cluster

Thanks to all the contributors involved in the development, such as @mendrugory, @bsctl, @maxgio92, and @alegrey91, @ptx96 although not directly involved in the Go code development! 🚀