v1.8.7 🌈

Release Notes for starrocks-kubernetes-operator v1.8.7

We are pleased to announce the release of starrocks-kubernetes-operator v1.8.7. This release includes several updates
and improvements that enhance the functionality and usability of the StarRocks Kubernetes Operator.

Enhancement

1. Added failure seconds for liveness and readiness. When StarRocks is under heavy load, the default Liveness Probe and
   Readiness Probe may fail, causing the container to restart. This update mitigates this issue by adding failure
   seconds for liveness and readiness.

Maintenance

1. [Documentation] Updated README.md and README_ZH-CN.md. This update provides more accurate and comprehensive
   information about StarRocks Operator.
2. [Documentation] Added local_installation_how_to.md for users. This guide provides step-by-step instructions for local installation
   of the Operator and StarRocks Cluster. And it also provides a script to help users install the Operator and StarRocks Cluster locally.

We encourage you to update to this new version and benefit from these improvements. As always, your feedback is very
welcome.

Changes

• [Chore] change related fields to release v1.8.7 @yandongxiao (#312)
• [Documentation] update README_ZH-CN.md @yandongxiao (#311)
• [Documentation] update README.md @yandongxiao (#310)
• [Enhancement] add failure seconds for liveness and readiness @yandongxiao (#309)
• [Documentation] add local installation how to.md for chinese users @yandongxiao (#307)
• [Chore] Improve helm repo update command @yandongxiao (#306)

v1.8.6 🌈

Release Notes for starrocks-kubernetes-operator v1.8.6

We are pleased to announce the release of starrocks-kubernetes-operator v1.8.6. This release includes several bug fixes
and enhancements. Here are the key updates:

Bug Fixes

1. Fix the problem of nginx sending the request body to FE when redirecting the stream load request, which may cause
   the stream load to fail. (#303)

Maintenance

1. [Documentation] add doc/load_data_using_stream_load.md. This document introduces how to load data from outside the
   k8s network to StarRocks through fe-proxy.
2. [Documentation] update change_root_password_howto.md. This document adds the steps of how to update the
   root password through Helm Chart.
3. [Chore] Add GitHub Actions to add label on issue and PR. This chore improves the project quality by adding necessary
   labels to issues and PRs.

We encourage you to update to this new version and benefit from these improvements. As always, your feedback is very
welcome.

PRS

• [Doc] add load_data_using_stream_load.md @yandongxiao (#296)
• [Doc] update change_root_password_howto.md @yandongxiao (#295)
• [Bugfix] set proxy_pass_request_body off when redirect to FE @yandongxiao (#303)
• [Documentation] update automatic_scaling_for_cn_nodes_howto.md @yandongxiao (#299)
• [Chore] Add Github Actions to add label on issue and PR @yandongxiao (#297)

v1.8.5 🌈

Release Notes for starrocks-kubernetes-operator v1.8.5

We are pleased to announce the release of starrocks-kubernetes-operator v1.8.5. This release includes several bug fixes
and enhancements. Here are the key updates:

What's New

1. [Chart] Service Account Customization: The operator now supports to provide the ability to add custom
   annotations and labels. The previous operator.global.rbac.serviceAccountName field is no longer supported.
2. [Operator] Explicit Protocol Selection for Istio: To provide additional functionality such as routing and rich
   metrics, Istio needs to determine the protocol of the traffic. This is now achieved by automatically adding the
   appProtocol field. This enhancement is particularly important for "server-first" protocols like MySQL, which are
   incompatible with automatic protocol detection and may cause connection failures.

Bug Fixes

1. Fixed an issue that occurred when starrocks.initPassword.enabled is true and the value
   of starrocks.starrocksCluster.name is set. The FE service domain name would follow the set value, while the initpwd
   pod would still use the starrocks.nameOverride value to compose the FE service domain name. This fix ensures
   consistent hostname usage.

Notes for Deployed Users

If starrocks.starrocksCluster.name is not set, the result of helm template will remain the same as before. If
starrocks.starrocksCluster.name is set and is different from the value of starrocks.nameOverride, the old configmaps
for FE, BE, and CN will be deleted. New configmaps with the new name for FE, BE, and CN will be created. This may
result in the restart of FE/BE/CN pods.

We encourage you to update to this new version and benefit from these improvements. As always, your feedback is very
welcome.

Changes

• [Bugfix] remove chart function kube-starrocks.name @yandongxiao (#292)

🚀 Features

• [Enhancement] add appProtocol in fe service @yandongxiao (#288)
• [Enhancement] create serviceaccount by default, and add ability to provide custom a… @AdamStaudt (#291)

v1.8.4 🌈

Release Notes for starrocks-kubernetes-operator v1.8.4

We are excited to announce the release of starrocks-kubernetes-operator v1.8.4, a Kubernetes operator for StarRocks.
This release brings several new features, bug fixes, and documentation improvements.

What's New

1. [Feature] add golangci-lint GitHub Action: This feature enables the golangci-lint tool to check the code quality
   and style of the operator project on every pull request.
2. [Feature] allow you to configure the terminationGracePeriodSeconds in StarRocksCluster CRD: This feature allows
   you to specify how long to wait before forcefully terminating a pod when deleting or updating a StarRocksCluster
   resource.
3. [Feature] add storage fields for starrocksCnSpec in values.yaml: This feature allows you to configure the storage
   class and size for the CN nodes in the StarRocksCluster.
4. [Feature] integration with Prometheus by ServiceMonitor CRD: This feature allows you to monitor the metrics of
   StarRocks cluster by using Prometheus and ServiceMonitor CRD.
5. [Feature] support startupProbeFailureSeconds fields in StarRocksCluster CRD: This feature allows you to configure
   the startup probe failure threshold for the pods in the StarRocksCluster resource.
6. [Feature] facilitate the configuration of environmental variables for the operator pod: for instance, one may
   designate KUBE_STARROCKS_UNSUPPORTED_ENVS to eliminate environments incompatible with the kubernetes cluster.

Bug Fixes

1. [Bugfix] This bugfix solves the problem that FE Proxy cannot handle stream load requests correctly when there are
   multiple FE pods in the StarRocks cluster.

Maintenance

1. [Doc] Set up StarRocks locally: This document provides a guide on how to set up a local StarRocks cluster.
2. [Doc] Establish a comprehensive StarRocks cluster, encompassing all available features. For additional examples
   concerning StarRocksCluster,
   please refer to: https://github.com/StarRocks/starrocks-kubernetes-operator/tree/main/examples/starrocks.
3. [Doc] Augment the instructional material on StarRocks utilization, encompassing topics such as '
   logging_and_related_configurations_howto.md' and 'mount_external_configmaps_or_secrets_howto.md'.
   For a holistic view of available guides,
   please refer to: https://github.com/StarRocks/starrocks-kubernetes-operator/tree/main/doc

Changes

🚀 Features

• [Feature] add golangci-lint Action @yandongxiao (#285)
• [Feature] allow to config the terminationGracePeriodSeconds @yandongxiao (#283)
• [Feature] add storage spec fields for starrocks cn spec in values.yaml @yandongxiao (#280)
• [Feature] integration with prometheus @yandongxiao (#284)
• [Feature] support customize probe @yandongxiao (#271)

🐛 Bug Fixes

• [Bugfix] fe proxy failed in multiple fe pods @yandongxiao (#269)
• [Bugfix] support k8s env which has no status.hostIP field @dujl (#266)

🧰 Maintenance

• [Doc] change related fields to release v1.8.4 @yandongxiao (#286)
• [Docs] install starrocks locally @yandongxiao (#274)
• [Doc] update examples/starrocks/deploy_a_starrocks_cluster_with_all_f… @yandongxiao (#278)
• [Doc] update doc/mount_external_configmaps_or_secrets_howto.md @yandongxiao (#277)
• [Docs] add starrocks-fe-and-be.yaml @yandongxiao (#276)
• [Doc] StarRocks Logging and Related Configurations @yandongxiao (#267)
• [Doc] Update deploy_a_starrocks_cluster_with_share_data_mode.yaml @DanRoscigno (#273)
• [Refactor] rename some name in pkg/k8sutils/templates/pod/spec.go @yandongxiao (#272)
• [Doc] add examples for StarRocksCluster CRD @yandongxiao (#270)
• [Doc] update comments in values.yaml @yandongxiao (#265)

v1.8.3 🌈

This is a minor release of StarRocks Kubernetes Operator.

What's New

1. [operator] Set proxy_read_timeout 600s: We have set proxy_read_timeout to 600s in nginx.conf, because the default value is 60s, which may cause timeout.
2. [chart] Add JAVA_OPTS_FOR_JDK_11 in FE config: We have added JAVA_OPTS_FOR_JDK_11 in FE config, because the default value is not suitable for JDK 11. If you used the default config of FE in values.yaml, Upgrading to v1.8.3 will cause the pods of FE to restart.
3. [chart] Allow user to specify serviceAccount for operator: By default the operator chart will create a serviceAccount for the operator, named starrocks. But if you want to use an existing serviceAccount, you can specify it in values.yaml.
4. [operator] Support Ports for feProxy component: We have supported Ports for feProxy component, allowing users to specify the nodePort of feProxy service.
5. [operator] Add namespace flag: It makes operator watch resources in a single namespace instead of all namespaces in the cluster. In most cases, you should not set this value. If your kubernetes cluster manages too many nodes, and operator watching all namespaces use too many memory resources, you can set this value.

Changes

• change related docs to release v1.8.3 @yandongxiao (#264)
• [Bugfix] change set proxy_read_timeout in nginx.conf @yandongxiao (#259)
• [Docs] Add JAVA_OPTS_FOR_JDK_11 to values.yaml @yandongxiao (#257)

🚀 Features

• add namespace flag to make operator watch resources in a single namespace @yandongxiao (#261)
• allow user to specify serviceAccount for operator @yandongxiao (#262)
• [Feature] Support nodePort for feProxy component @yandongxiao (#258)

v1.8.2 🌈

This is a minor release of StarRocks Kubernetes Operator, a project that aims to provide a Kubernetes-native way to
deploy and manage StarRocks clusters on Kubernetes.

What's New

1. [operator.yaml] Change the resource of operator: We have changed the default resource limit and request of
   operator pod in case of OOM.
2. [operator] Remove the watch for the HPA resource: We have removed the watch for the HPA resource, because
   operator can not make sure what version of HPA is used in the cluster.

Changes

• [Docs] change related fields to release v1.8.2 @yandongxiao (#255)
• [Bugfix] update operator memory resouce @yandongxiao (#254)
• [Bugfix] unable to get k8s version when start operator by host node @yandongxiao (#253)
• [Bugfix] Remove the watching for v2 HPA @yandongxiao (#252)

v1.8.1 🌈

We are excited to announce the release of StarRocks Kubernetes Operator v1.8.1. This release brings some new features and bug fixes to improve the user experience of deploying and managing StarRocks clusters on Kubernetes.

What's New

1. [operator] Watch HPA: The operator will watch the HPA resources and once the HPA is missing, the operator will
   try to create it again.
2. [operator] Select the appropriate HPA version: The operator will select the appropriate HPA version based on the
   Kubernetes version, ensuring compatibility and stability.
3. [operator] Reduce incorrect modification of replicas: The operator will remove the replicas field of the CN
   statefulset when HPA is enabled, avoiding conflicts between HPA and statefulset controller.
4. [operator] Support subpath: The operator will support subpath for configmaps and secrets, allowing users to mount
   specific files or directories from these resources.
5. [operator] Remove unnecessary resources: The operator will remove the related Kubernetes resources when the
   BeSpec or CnSpec of the StarRocks cluster is deleted, ensuring a clean and consistent state of the cluster.
6. [operator] Support ports fields: The operator will support ports field for the StarRocks cluster, allowing users
   to customize the ports of the services.

Changes

• [Docs] change related fields to release v1.8.1 @yandongxiao (#250)
• Bugfix/watch hpa resources @yandongxiao (#246)
• update documents @yandongxiao (#241)
• remove service.name fields @yandongxiao (#243)

🚀 Features

• Feature/select appropriate hpa version @yandongxiao (#248)
• remove replicas fields of CN Statefulset when hpa is enabled @yandongxiao (#247)
• support subpath for configmaps and secrets @yandongxiao (#249)
• remove related k8s resources when BeSpec/CnSpec is deleted @yandongxiao (#245)
• support ports fields @yandongxiao (#244)
• remove restart logic @yandongxiao (#242)
• update some scripts for v1.8.0 @yandongxiao (#237)

v1.8.0 🌈

We are excited to announce the release of StarRocks Kubernetes Operator v1.8.0. This release brings some new features and bug fixes to improve the user experience of deploying and managing StarRocks clusters on Kubernetes.

What's New

• [operator] CRD changes: We have modified the CRD definition in v1.8.0. Before you use helm upgrade to upgrade from a previous version, you need to manually update the CRD definition.
• [operator] Non-root mode: We have added a runAsNonRoot field to support running StarRocks pods as non-root users. This can enhance the security of your StarRocks clusters.
• [chart] Subchart support: We have split the kube-starrocks chart into two subcharts: operator and starrocks. Installing kube-starrocks is equivalent to installing both operator and starrocks subcharts, and uninstalling kube-starrocks is equivalent to uninstalling both operator and starrocks subcharts. If you want more flexibility in managing your StarRocks clusters, you can install operator and starrocks subcharts separately.
• [chart] Values migration tool: We have provided a tool to migrate your values.yaml file from a previous version to v1.8.0. You can use the migrate-chart-value command to upgrade your values.yaml file.
• [chart] Multiple cluster support: We have enabled support for deploying multiple StarRocks clusters in different namespaces in one Kubernetes cluster. You just need to install the starrocks chart in different namespaces.
• [operator] FE proxy feature: We have added the FE proxy feature to allow external clients and data import tools to access StarRocks clusters in Kubernetes. For example, you can use the STREAM LOAD syntax to import data into StarRocks clusters.
• [chart] Datadog integration: We have integrated with Datadog to provide metrics and logs for StarRocks clusters. You can enable this feature by setting the datadog related fields in your values.yaml file.
• [chart] Cluster password initialization: We have added the ability to initialize the password of root in your StarRocks cluster during installation. Note that this only works for helm install, can't use it in helm upgrade

How to Upgrade

To upgrade from a previous version of StarRocks Kubernetes Operator, please follow these steps:

1. Update the CRD definition by running: kubectl apply -f https://github.com/StarRocks/starrocks-kubernetes-operator/releases/download/v1.8.0/starrocks.com_starrocksclusters.yaml
2. Migrate your values.yaml file by running: migrate-chart-value --input values.yaml --target-version v1.8.0 --output ./values-v1.8.0.yaml
3. Upgrade your Helm release by running: helm repo update; helm upgrade <release-name> starrocks-community/kube-starrocks -f values-v1.8.0.yaml

How to Install

To install StarRocks Kubernetes Operator for the first time, please follow these steps:

1. Add the StarRocks Helm repository by running: helm repo add starrocks-community https://starrocks.github.io/starrocks-kubernetes-operator, then execute helm repo update.
2. Install the kube-starrocks chart by running: helm install <release-name> starrocks-community/kube-starrocks -f values.yaml
3. Alternatively, you can install the operator and starrocks subcharts separately by running. Install operator: helm install <release-name> starrocks-community/operator -f operator-values.yaml. Install starrocks: helm install <starrocks-release-name> starrocks-community/starrocks -f starrocks-values.yaml

Feedback

We welcome your feedback and contributions to StarRocks Kubernetes Operator. If you have any questions, suggestions, or issues, please feel free to open an issue on our GitHub repository. Thank you for using StarRocks Kubernetes Operator!

Changes

• add ResourceVersion before update service resources @yandongxiao (#217)
• [refactor] rename some names @yandongxiao (#207)
• doc/add prometheus grafana integration doc @kevincai (#202)
• [Bugfix] race condition on fe/be/cn status when upgrade @yandongxiao (#198)
• Add time zone settings and fixed issue where Secret could not find during helm upgrade. @WholeWorld-Timothy (#188)
• update index.yaml @yandongxiao (#190)

🚀 Features

• add fe proxy @yandongxiao (#211)
• set resync period @yandongxiao (#218)
• support subpath in crd @yandongxiao (#212)
• change the default volume size of fe meta @yandongxiao (#210)
• allow user to add attributes on datadog log annotations @yandongxiao (#208)
• add tool to upgrade or downgrade values.yaml file @yandongxiao (#206)
• Feature/use helm template to generate operator yaml file @yandongxiao (#200)
• Feature/use subchart to support multiple starrocks cluster @yandongxiao (#199)
• reformat go imports @yandongxiao (#201)
• feature/integration-with-datadog @yandongxiao (#197)
• [Feature] add runAsNonRoot filed @yandongxiao (#195)
• Bugfix/remove version from selector @yandongxiao (#191)

v1.7.1 🌈

Notes:

1. operator will reconcile when service type updated.
2. init password job will only be executed when helm install
3. do some work about artifacthub.io
4. add some security configurations on pod.

Changes

• [doc] add namespace back in example yaml @kevincai (#181)

🚀 Features

• feature run as non root @yandongxiao (#184)
• Publish PGP public key so that tools can verify the chart @yandongxiao (#180)
• add provenance and integrity for helm chart @yandongxiao (#178)
• [doc] add artifacthub-repo.yml and update README.md @yandongxiao (#177)

🐛 Bug Fixes

• feature: remove cleanup logic when starrocks cluster is deleted @yandongxiao (#187)
• bugfix/add service type when calculate hash @yandongxiao (#186)
• Modify the execution time of the job to only execute the job when executing 'helm install'. @WholeWorld-Timothy (#185)
• use brackets to change priority when inject environment @yandongxiao (#182)

v1.7.0 🌈

Notes

• In this new release, POD probe is switched from TCP/9030 to HTTP/8030, as a result, existing Pods (FE/BE/CN) will be rolling restarted after the operator upgraded.

Changes

• move index yaml to root @yandongxiao (#175)
• Add instructions for initializing password configuration methods @WholeWorld-Timothy (#171)
• improve github workflows @yandongxiao (#168)
• symlink helm charts crd definition to the operator's one @kevincai (#119)
• doc: Update change_root_password_howto.md @jzcruiser (#123)
• Merge helm charts into the operator repo @dengliu (#117)
• [DOC] add howto doc with instructions to change root password @kevincai (#111)

🚀 Features

• Feature/add test cases @yandongxiao (#173)
• [feature] add release-drafter and release-drafter actions, and add artifacts.sh script @yandongxiao (#167)
• Feature/support to config toleration node selector annotation for operator pod @yandongxiao (#165)
• [feature] make command make work, and add github workflows @yandongxiao (#162)
• Feature/ensure pods restart when configmap changes @yandongxiao (#157)
• [feature] support volumes mounting for configmaps @yandongxiao (#156)
• 在部署starrocks时，增加初始化密码的能力。 @WholeWorld-Timothy (#133)
• [feature] support to add annotations on k8s service @yandongxiao (#139)
• [feature] support customized pod scheduler @yandongxiao (#137)
• [feature] remove GOARCH flag to support build multi-arch image @yandongxiao (#135)
• [Enhancement] Use HTTP endpoint for StartupProbe/Liveness/Readiness P… @yandongxiao (#132)
• [feature] support mount secrets in operator and helm chart @yandongxiao (#129)
• 增加logStorageSize配置,当增加此配置时,fe和be的log存储使用pvc管理 @WholeWorld-Timothy (#124)

🐛 Bug Fixes

• [bugfix] fix some bugs @yandongxiao (#172)
• bugfix service ownerReferences @yandongxiao (#170)
• [bugfix] if feEnvVars has values and initPassword.enabled, helm template will fail @yandongxiao (#169)
• [bugfix] remove finalizers for existing cluster @yandongxiao (#150)
• fix wrong fe service labels @kevincai (#98)
• [bugfix] make commands make run && make install work @yandongxiao (#127)
• fix log positional parameters @kevincai (#97)
• fix crd yaml too long to apply by kubectl @kevincai (#99)

🧰 Maintenance

• cleanup redundant yaml files in deploy directory @yandongxiao (#164)
• add vendor @yandongxiao (#163)
• Refactor/add a leading space to comments @yandongxiao (#158)
• [refactor] refactoring code to remove duplicate code @yandongxiao (#152)
• [doc] add license for go files @yandongxiao (#151)
• update image tag in examples yaml @kevincai (#149)
• remove finalizer and refactor some parts to remove some duplicate code @yandongxiao (#148)
• clean up helm chart values.yaml @yandongxiao (#141)
• use command make fmt to reformat code @yandongxiao (#138)
• preserve configmap eol for fe/be/cn @kevincai (#134)
• [refactor] change some log format @yandongxiao (#128)