v3.0.0

TKS - Release v3.0.0

TKS는 SK텔레콤에서 개발한 GitOps, Declarative 방식의 Kubernetes 관리 및 다양한 앱현대화 서비스를 제공하는 플랫폼입니다. SKT에서는 TKS 소프트웨어를 기반으로 1) TKS 클라우드서비스 (SaaS)와 2) TKS 엔터프라이즈솔루션 (구축형상품)을 출시하여 제공하고 있습니다.

본 v3.0.0 릴리즈에는 Kustomize, Helm, Argo Workflow등을 활용하여 복잡한 소프트웨어들을 논리적으로 그룹핑하여 단순화시키고 이를 선언적으로 관리하기 위한 기술인 Decapod를 기반으로 하여, 인프라, Kubernetes플랫폼, 그리고 플랫폼 운영 및 앱현대화를 위해 필요한 소프트웨어 세트들 (LMA, MSA)을 일괄 자동 배포/관리하는 "스택" 기능이 추가되었습니다. 또한, 모든 기능들은 API로 제공됩니다.

이번 릴리즈에서 추가된 신규 기능들은 다음과 같습니다.

1. AWS의 EKS연동 및 On-prem/Private Cloud 지원을 위한 BYOH (Bring Your Own Kubernetes) 프로바이더 지원기능이 추가되었습니다.
2. LMA (Logging, Monitoring, Alerting) 세트에서 Elastic Search와 Kibana가 Deprecated되고, Grafana Loki와 Thanos가 신규로 도입되었습니다. 또한, Thanos, Loki를 기반으로 한 멀티클러스터 통합 모니터링 체계를 Reference Archiecture로 제공합니다.
3. Service Mesh에서 Tracing Log 저장을 위한 Elastic Search가 Deprecated 되고, Cassandra가 신규 도입되었습니다.
4. GitOps를 위해 Gitea와 Harbor가 기본 저장소로 제공됩니다.
5. 멀티테넌시를 위한 Organization이 최초로 적용되었고, Organization별 사용자 등록, 관리가 가능합니다.
6. SaaS형 서비스인 TKS 클라우드서비스를 위한 Admin, 사용자 포탈 UI가 신규로 제공됩니다.
7. 이 외에 많은 버그수정과 안정화 작업을 진행하였습니다.

감사합니다.

What's Changed

• change thanos service type to LoadBalancer by @robertchoi80 in #159
• disable dns discovery of thanos sidecar service by @robertchoi80 in #160
• merge custom-base-yaml by @robertchoi80 in #161
• lma: use downstream charts for thanos and kube-state-metrics by @zugwan in #162
• remove duplicate key by @seungkyua in #163
• support multiple infra provider by @zugwan in #164
• add tks-cluster-byoh group by @zugwan in #165
• install local-path-provioner for a byoh cluster by @intelliguy in #166
• using base-overlay to support multi-infra by @intelliguy in #167
• lma.fluentbit: support multiline and multi-es by @intelliguy in #168
• fluentbit의 기본값을 TACO LMA형상에 맞도록 수정 by @intelliguy in #169
• TACO LMA 로깅의 백엔드 선택기능 추가 by @intelliguy in #170
• bugfix: escape from kustomize error by @intelliguy in #171
• fluentbit crd변경 반영 by @intelliguy in #172
• Loki production by @intelliguy in #173
• minio: dedicated installation for shared service(thanos/loki) by @intelliguy in #174
• minio: configration for dedicated installation for shared service(tha… by @intelliguy in #175
• Update resources.yaml by @ktkfree in #177
• fluentbit: change the default loki target by @intelliguy in #178
• lma-addons: new mechanism for deploing grafana dashboard by @intelliguy in #176
• bugfix. disable kibana-init by @ktkfree in #179
• bump up lma-addons by @intelliguy in #180
• loki: add nginx config to recieve massive data by @intelliguy in #181
• add jaeger nodeAffinity by @seungkyua in #182
• add jaeger image and istio-ingress-gateway image url by @seungkyua in #183
• add pilot image url by @seungkyua in #184
• event-exporter: add a way to configure cluster-name by @intelliguy in #185
• clustar-api-aws 정적 워커 노드 배포 by @zugwan in #186
• update cluster-api-aws to 0.8.0 for EKS by @zugwan in #190
• update cluster-api-aws for eks addons by @zugwan in #192
• change original url to tks harbor url by @seungkyua in #195
• feature: Change minio values to support Arm64 platform openinfradev/tks-issues#500 by @cho4036 in #187
• change default for single cluster by @intelliguy in #191
• 지정된 저장소를 사용하는지 CI단계에서 체크하는 로직 추가 by @intelliguy in #189
• lma: remove prometheus config reloader tag by @zugwan in #196
• minio: Reduce default request memory size 16->1GB by @cho4036 in #197
• use harbor for container image registry by @zugwan in #198
• update local chart repository for check by @intelliguy in #199
• remove unused charts by @intelliguy in #200
• change lma images to use local by @intelliguy in #201
• upgrade charts with psp to support kube v1.25 by @intelliguy in #194
• lma: fix prometheus-operator images by @zugwan in #203
• fix prometheus and alertmanager instance image by @zugwan in #204
• lma: user helm chart's default tag for kube-state-metrics image by @zugwan in #205
• 원본 Helm 이력 추적을 위한 url 추가 by @cho4036 in #202
• Fix missing harbor images by @zugwan in #206
• cluster-api-aws: change default POD CIDR by @zugwan in #207
• ingress-nginx: remove image digest by @zugwan in #208
• add ack for s3 by @intelliguy in #209
• primary cluster 체계를 위한 기본값 변경 by @intelliguy in #210
• update cluster-api-aws to 0.8.2 for multi-tenancy by @zugwan in #211
• tks-cluster: ingress-nginx: remove unnecessary service port by @zugwan in #213
• upgrade msa by @seungkyua in #214
• add cassandra chart by @seungkyua in #215
• fix nodeSelector by @seungkyua in #216
• fix nodeSelector by @seungkyua in #217
• add k8ssandra tool image by @seungkyua in #218
• tks-cluster: aws: update k8s version to v1.25.9 by @zugwan in #219
• bump kube-state-metrics up to v3.3.4 by @intelliguy in #220
• add argo-rollouts chart by @seungkyua in #222
• fix indent by @seungkyua in #223
• typo in alermanager configuration. by @intelliguy in #224
• ACK: update bucket name by @intelliguy in #212
• alert용 slack channel 지정 by @intelliguy in #225
• alertmanager: remove a route for slack alert by @intelliguy in #226
• add upstream cluster-autoscaler chart by @zugwan in #227
• grafana: add node selector by @intelliguy in #228
• tks-cluster: k8s-cluster-autoscaler: fix value indent by @zugwan in #229
• tks-cluster: k8s-cluster-autoscaler: make value to string by @zugwan in #230
• prometheus: override alertmanager config for multiple routes by @intelliguy in #231
• tks-cluster: aws: use cluster-api-aws chart v0.8.3 by @zugwan in #234
• fluent-resource: using fluentbit agent - v2.1.4 by @intelliguy in #235
• minio: add default ilm by @intelliguy in #233
• prepare-etcd-secret: apply the label changes after v1.20 by @intelliguy in #236
• minio: bugfix for value level by @intelliguy in #237
• typo: fluentbit: remove garbage when define image by @intelliguy in #238
• minio: change deploy mode from standalone to distribute by @intelliguy in #239
• minio: change default replicas number to 1 by @intelliguy in #241
• minio: change update strategy to avoid competition for pvc by @intelliguy in #242
• S3support by @intelliguy in #243
• bumpup thanos-config by @intelliguy in #247
• bugfix: remove default values for minio by @intelliguy in #249
• ack-resources: add controllerability for versioned objects on s3 buckets by @intelliguy in #252
• release -> main merge (20220808) by @ktkfree in #253
• main merge ( 20230814 ) by @ktkfree in #255
• change image repo from harbor-cicd to harbor by @intelliguy in #259
• main merge (20230816) by @ktkfree in #260

New Contributors

• @cho4036 made their first contribution in #187

Full Changelog: v2.0.0...v3.0.0