K8S 离线安装 dolphinscheduler2.0.5
由于环境原因，不能在生产环境访问外网，所以只能在外网下载好DS相关的依赖镜像上传到内网环境，在内网环境安装，k8s部署流程参考官方文档：快速试用 Kubernetes 部署。
一、下载镜像到本地
[root@hombd03 dolphinscheduler] cd /opt/modules/apache-dolphinscheduler-2.0.5-src/docker/kubernetes/dolphinscheduler在 values.yaml 文件中修改镜像，以实现离线安装（air-gap install）；
image:
repository: "apache/dolphinscheduler"
tag: "2.0.5"
pullPolicy: "IfNotPresent"
pullSecret: ""针对公司内部安装好的harbor，或者其他公有云的私仓，进行pull，tag，以及push。这里我们假定私仓地址是harbor.abc.com，你所在构建镜像的主机已经进行了docker login harbor.abc.com， 且已经建立和授权私仓下面新建apache项目。
执行shell命令：
# 下载镜像到本地
docker pull apache/dolphinscheduler:2.0.5
# 将下载的镜像打成tar包，然后导入到内网（该内网环境不能访问外网）
dock tag apache/dolphinscheduler:2.0.5 harbor.abc.com/apache/dolphinscheduler:2.0.5
docker push apache/dolphinscheduler:2.0.5再替换value文件中的镜像信息，这里我们推荐使用Always的方式拉取镜像，生产环境中尽量每次去检查是否是最新的镜像内容，保证软件制品的正确性。此外，很多同学会把tag写成latest（制作镜像不写tag信息，这样在生产环境非常危险，任何人push了镜像，就相当于改变了latest的tag的镜像，而且也无法判断latest是什么版本，所以建议要明确每次发版的tag，并且使用Always。
image:
repository: "harbor.abc.com/apache/dolphinscheduler"
tag: "2.0.5"
pullPolicy: "Always"更新DS依赖
$ helm repo add bitnami https://charts.bitnami.com/bitnami
$ helm dependency update .报如下的错，postgresql:10.3.18, zookeeper:6.5.3 版本找不到对应的包：
[root@homaybd03 dolphinscheduler]# helm dependency update .
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "bitnami" chart repository
Update Complete. ⎈Happy Helming!⎈
Error: can't get a valid version for repositories postgresql, zookeeper. Try changing the version constraint in Chart.yaml如果找不到对应的包，则通过 helm search repo postgresql -l 命令来搜索可使用的包：
[root@homaybd03 dolphinscheduler]# helm search repo postgresql -l
NAME CHART VERSION APP VERSION DESCRIPTION
bitnami/postgresql 11.6.13 14.4.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.6.12 14.4.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.6.11 14.4.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.6.10 14.4.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.6.9 14.4.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.6.8 14.4.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.6.7 14.4.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.6.6 14.3.0 PostgreSQL (Postgres) is an open source object-...
...
bitnami/postgresql 11.1.13 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.12 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.11 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.10 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.9 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.8 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.7 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.6 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.5 14.2.0 PostgreSQL (Postgres) is an open source object-...
bitnami/postgresql 11.1.4 14.2.0 PostgreSQL (Postgres) is an open source object-...
...
bitnami/postgresql 10.14.4 11.14.0 Chart for PostgreSQL, an object-relational data...
bitnami/postgresql 10.14.3 11.14.0 Chart for PostgreSQL, an object-relational data...
bitnami/postgresql-ha 9.2.0 14.4.0 This PostgreSQL cluster solution includes the P...
bitnami/postgresql-ha 9.1.9 14.4.0 This PostgreSQL cluster solution includes the P...
bitnami/postgresql-ha 9.1.8 14.4.0 This PostgreSQL cluster solution includes the P...
bitnami/postgresql-ha 9.1.7 14.4.0 This PostgreSQL cluster solution includes the P...
bitnami/postgresql-ha 9.1.6 14.4.0 This PostgreSQL cluster solution includes the P...
bitnami/postgresql-ha 9.1.5 14.3.0 This PostgreSQL cluster solution includes the P...
bitnami/postgresql-ha 9.1.4 14.3.0 This PostgreSQL cluster solution includes the P...
bitnami/postgresql-ha 9.1.2 14.3.0 This PostgreSQL cluster solution includes the P...
bitnami/postgresql-ha 9.1.1 14.3.0 This PostgreSQL cluster solution includes the P...
...
bitnami/postgresql-ha 8.2.1 11.14.0 Chart for PostgreSQL with HA architecture (usin...
bitnami/postgresql-ha 8.2.0 11.14.0 Chart for PostgreSQL with HA architecture (usin...
bitnami/postgresql-ha 8.1.5 11.14.0 Chart for PostgreSQL with HA architecture (usin...
[root@homaybd03 dolphinscheduler]# 查看zookeepr可用的版本：
[root@homaybd03 dolphinscheduler]# helm search repo zookeeper -l
NAME CHART VERSION APP VERSION DESCRIPTION
bitnami/zookeeper 10.0.1 3.8.0 Apache ZooKeeper provides a reliable, centraliz...
bitnami/zookeeper 10.0.0 3.8.0 Apache ZooKeeper provides a reliable, centraliz...
bitnami/zookeeper 9.2.7 3.8.0 Apache ZooKeeper provides a reliable, centraliz...
bitnami/zookeeper 9.2.6 3.8.0 Apache ZooKeeper provides a reliable, centraliz...
...
bitnami/zookeeper 8.0.3 3.7.0 Apache ZooKeeper provides a reliable, centraliz...
bitnami/zookeeper 8.0.2 3.7.0 Apache ZooKeeper provides a reliable, centraliz...
bitnami/zookeeper 8.0.1 3.7.0 A centralized service for maintaining configura...
bitnami/zookeeper 8.0.0 3.7.0 A centralized service for maintaining configura...
bitnami/zookeeper 7.6.2 3.7.0 A centralized service for maintaining configura...
bitnami/zookeeper 7.6.1 3.7.0 A centralized service for maintaining configura...
bitnami/zookeeper 7.6.0 3.7.0 A centralized service for maintaining configura...
...
bitnami/dataplatform-bp2 10.0.10 1.0.1 This Helm chart can be used for the automated d...
bitnami/dataplatform-bp2 10.0.9 1.0.1 This Helm chart can be used for the automated d...
bitnami/dataplatform-bp2 10.0.8 1.0.1 OCTO Data platform Kafka-Spark-Elasticsearch He...
bitnami/kafka 16.2.5 3.1.0 Apache Kafka is a distributed streaming platfor...
bitnami/kafka 16.2.4 3.1.0 Apache Kafka is a distributed streaming platfor...
bitnami/kafka 16.2.3 3.1.0 Apache Kafka is a distributed streaming platfor...
...
bitnami/kafka 15.3.5 3.1.0 Apache Kafka is a distributed streaming platfor...
bitnami/solr 4.1.0 8.11.1 Apache Solr is an extremely powerful, open sour...
bitnami/solr 4.0.12 8.11.1 Apache Solr is an extremely powerful, open sour...
bitnami/solr 4.0.11 8.11.1 Apache Solr is an extremely powerful, open sour...
bitnami/solr 4.0.10 8.11.1 Apache Solr is an extremely powerful, open sour...
bitnami/solr 4.0.9 8.11.1 Apache Solr is an extremely powerful, open sour...
bitnami/solr 4.0.8 8.11.1 Apache Solr is an extremely powerful, open sour...
bitnami/solr 4.0.7 8.11.1 Apache Solr is an extremely powerful, open sour...
...
bitnami/solr 2.4.2 8.11.1 Apache Solr is an open source enterprise search...
bitnami/solr 2.3.0 8.11.1 Apache Solr is an open source enterprise search...
bitnami/solr 2.2.0 8.11.1 Apache Solr is an open source enterprise search...
[root@homaybd03 dolphinscheduler]# 更改 Chart.yaml 文件的版本：
dependencies:
- name: postgresql
version: 10.14.3
# Due to a change in the Bitnami repo, https://charts.bitnami.com/bitnami was truncated only
# containing entries for the latest 6 months (from January 2022 on).
# This URL: https://raw.githubusercontent.com/bitnami/charts/archive-full-index/bitnami
# contains the full 'index.yaml'.
# See detail here: https://github.com/bitnami/charts/issues/10833
repository: https://raw.githubusercontent.com/bitnami/charts/archive-full-index/bitnami
condition: postgresql.enabled
- name: zookeeper
version: 8.0.2
# Same as above.
repository: https://raw.githubusercontent.com/bitnami/charts/archive-full-index/bitnami
condition: zookeeper.enabled然后再重新加载：
$ helm repo add bitnami https://charts.bitnami.com/bitnami
$ helm dependency update .安装DS：
# [root@hombd03 dolphinscheduler]# kubectl create ns test
helm install dolphinscheduler . --set image.tag=2.0.5 -n test将名为 dolphinscheduler 的版本(release) 发布到 test 的命名空间中：
[root@hombd03 dolphinscheduler]# helm install dolphinscheduler . --set image.tag=2.0.5 -n test
NAME: dolphinscheduler
LAST DEPLOYED: Tue Jul 5 00:08:05 2022
NAMESPACE: test
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
#
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
** Please be patient while the chart DolphinScheduler 2.0.5 is being deployed **
Access DolphinScheduler UI URL by:
kubectl port-forward -n test svc/dolphinscheduler-api 12345:12345
DolphinScheduler UI URL: http://127.0.0.1:12345/dolphinscheduler
[root@homaybd03 dolphinscheduler]# 查看Pod启动状况：
[root@hombd03 dolphinscheduler]# kubectl get pods -n test
NAME READY STATUS RESTARTS AGE
dolphinscheduler-alert-69bb5f4f95-w2dft 0/1 Running 0 2m52s
dolphinscheduler-api-84889f7f64-b9mzp 0/1 ContainerCreating 0 2m52s
dolphinscheduler-master-0 0/1 ContainerCreating 0 2m52s
dolphinscheduler-master-1 0/1 ContainerCreating 0 2m52s
dolphinscheduler-master-2 0/1 Running 0 2m52s
dolphinscheduler-postgresql-0 1/1 Running 0 2m52s
dolphinscheduler-worker-0 0/1 Running 0 2m52s
dolphinscheduler-worker-1 0/1 Running 0 2m52s
dolphinscheduler-worker-2 0/1 ContainerCreating 0 2m52s
dolphinscheduler-zookeeper-0 1/1 Running 0 2m52s
[root@hombd03 dolphinscheduler]# kubectl get pods -n test -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
dolphinscheduler-alert-69bb5f4f95-w2dft 1/1 Running 0 3m22s 192.200.72.222 hombd05 <none> <none>
dolphinscheduler-api-84889f7f64-b9mzp 0/1 Running 0 3m22s 192.200.245.154 hombd04 <none> <none>
dolphinscheduler-master-0 0/1 Running 0 3m22s 192.200.245.150 hombd04 <none> <none>
dolphinscheduler-master-1 0/1 Running 0 3m22s 192.200.245.153 hombd04 <none> <none>
dolphinscheduler-master-2 0/1 Running 0 3m22s 192.200.72.226 hombd05 <none> <none>
dolphinscheduler-postgresql-0 1/1 Running 0 3m22s 192.200.72.224 hombd05 <none> <none>
dolphinscheduler-worker-0 1/1 Running 0 3m22s 192.200.72.223 hombd05 <none> <none>
dolphinscheduler-worker-1 1/1 Running 0 3m22s 192.200.72.225 hombd05 <none> <none>
dolphinscheduler-worker-2 0/1 Running 0 3m22s 192.200.245.152 hombd04 <none> <none>
dolphinscheduler-zookeeper-0 1/1 Running 0 3m22s 192.200.245.151 hombd04 <none> <none>
[root@hombd03 dolphinscheduler]# 等一会在看pod启动：
[root@homaybd03 dolphinscheduler]# kubectl get pods -n test
NAME READY STATUS RESTARTS AGE
dolphinscheduler-alert-69bb5f4f95-w2dft 1/1 Running 0 27m
dolphinscheduler-api-84889f7f64-7vlnz 1/1 Running 0 4m53s
dolphinscheduler-master-0 1/1 Running 0 98s
dolphinscheduler-master-1 1/1 Running 0 97s
dolphinscheduler-master-2 1/1 Running 5 27m
dolphinscheduler-postgresql-0 1/1 Running 0 27m
dolphinscheduler-worker-0 1/1 Running 6 27m
dolphinscheduler-worker-1 0/1 CrashLoopBackOff 6 27m
dolphinscheduler-worker-2 1/1 Running 1 2m23s
dolphinscheduler-zookeeper-0 1/1 Running 0 100s
[root@homaybd03 dolphinscheduler]# kubectl get pods -n test -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
dolphinscheduler-alert-69bb5f4f95-w2dft 1/1 Running 0 27m 192.200.72.222 homaybd05 <none> <none>
dolphinscheduler-api-84889f7f64-7vlnz 1/1 Running 0 5m36s 192.200.72.233 homaybd05 <none> <none>
dolphinscheduler-master-0 1/1 Running 0 2m21s 192.200.245.161 homaybd04 <none> <none>
dolphinscheduler-master-1 1/1 Running 0 2m20s 192.200.245.162 homaybd04 <none> <none>
dolphinscheduler-master-2 1/1 Running 5 27m 192.200.72.226 homaybd05 <none> <none>
dolphinscheduler-postgresql-0 1/1 Running 0 27m 192.200.72.224 homaybd05 <none> <none>
dolphinscheduler-worker-0 1/1 Running 6 27m 192.200.72.223 homaybd05 <none> <none>
dolphinscheduler-worker-1 0/1 Running 7 27m 192.200.72.225 homaybd05 <none> <none>
dolphinscheduler-worker-2 1/1 Running 1 3m6s 192.200.245.155 homaybd04 <none> <none>
dolphinscheduler-zookeeper-0 1/1 Running 0 2m23s 192.200.245.159 homaybd04 <none> <none>
[root@homaybd03 dolphinscheduler]# 查看Pod日志：
[root@hombd03 dolphinscheduler]# kubectl logs dolphinscheduler-api-84889f7f64-b9mzp -n test
Error from server: Get "https://homaybd04:10250/containerLogs/test/dolphinscheduler-api-84889f7f64-b9mzp/dolphinscheduler-api": net/http: TLS handshake timeout
[root@homaybd03 dolphinscheduler]# 如果 values.yaml 文件中的 ingress.enabled 被设置为 true, 在浏览器中访问 http://${ingress.host}/dolphinscheduler 即可
提示: 如果 ingress 访问遇到问题，请联系 Kubernetes 管理员并查看 Ingress
否则，当 api.service.type=ClusterIP 时，你需要执行 port-forward 端口转发命令：
kubectl port-forward --address 0.0.0.0 -n test svc/dolphinscheduler-api 12345:12345 # 使用 test 命名空间[root@hombd03 dolphinscheduler]# kubectl get svc -n test
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
dolphinscheduler-alert ClusterIP 192.233.2.127 <none> 50052/TCP 34m
dolphinscheduler-api ClusterIP 192.233.214.70 <none> 12345/TCP 34m
dolphinscheduler-master-headless ClusterIP None <none> 5678/TCP 34m
dolphinscheduler-postgresql ClusterIP 192.233.60.228 <none> 5432/TCP 34m
dolphinscheduler-postgresql-headless ClusterIP None <none> 5432/TCP 34m
dolphinscheduler-worker-headless ClusterIP None <none> 1234/TCP,50051/TCP 34m
dolphinscheduler-zookeeper ClusterIP 192.233.241.93 <none> 2181/TCP,2888/TCP,3888/TCP 34m
dolphinscheduler-zookeeper-headless ClusterIP None <none> 2181/TCP,2888/TCP,3888/TCP 34m
[root@hombd03 dolphinscheduler]# 导出镜像包：
# https://blog.csdn.net/m0_67402235/article/details/124174771
[root@hombd03 saveimage]# docker save > dolphinscheduler-2.0.5.tar apache/dolphinscheduler:2.0.5
[root@hombd03 saveimage]# docker save > dolphinscheduler-mysql-2.0.5.tar apache/dolphinscheduler:mysql-driver
[root@hombd03 saveimage]# docker save > dolphinscheduler-scarf-2.0.5.tar dolphinscheduler.docker.scarf.sh/apache/dolphinscheduler
docker pull bitnami/zookeeper:3.7.0-debian-10-r295
docker save > zookeeper-3.7.0.tar bitnami/zookeeper:3.7.0-debian-10-r295导入镜像包：
# export 导出容器，对应的import导入
# docker import dolphinscheduler-mysql-2.0.5.tar apache/dolphinscheduler:2.0.5
# save对应的是load 使用 -i 导入
[root@dce88 ~]# docker load -i /opt/images.tar.gz
# 使用 < 导入
[root@dce88 ~]# docker load < /opt/images.tar.gzexport和save命令区别
export命令是从容器（container）中导出tar文件，而save命令则是从镜像（images）中导出
- 文件大小不同：export 导出的镜像文件体积小于 save 保存的镜像
- 是否可以对镜像重命名
docker import 可以为镜像指定新名称
docker load 不能对载入的镜像重命名 - 是否可以同时将多个镜像打包到一个文件中
docker export 不支持
docker save 支持 - 是否包含镜像历史
export 导出（import 导入）是根据容器拿到的镜像，再导入时会丢失镜像所有的历史记录和元数据信息（即仅保存容器当时的快照状态），所以无法进行回滚操作
而 save 保存（load 加载）的镜像，没有丢失镜像的历史，可以回滚到之前的层（layer）
应用场景不同docker export 的应用场景：主要用来制作基础镜像，比如我们从一个 ubuntu 镜像启动一个容器，然后安装一些软件和进行一些设置后，使用 docker export 保存为一个基础镜像。然后，把这个镜像分发给其他人使用，比如作为基础的开发环境。
docker save 的应用场景：如果我们的应用是使用 docker-compose.yml 编排的多个镜像组合，但我们要部署的客户服务器并不能连外网。这时就可以使用 docker save 将用到的镜像打个包，然后拷贝到客户服务器上使用 docker load 载入。
部署服务
启动zookeeper 服务，本地端口 8181 映射到容器的 2181 端口
docker run --privileged=true -d -e TZ="Asia/Shanghai" -e ALLOW_ANONYMOUS_LOGIN=yes -p 8181:2181 -v /data/zookeeper:/data --name zookeeper --restart always bitnami/zookeeper:3.7.0-debian-10-r295运行一个 DolphinScheduler 实例
$ docker run -d --name dolphinscheduler \
-e DATABASE_TYPE="mysql" -e DATABASE_DRIVER="com.mysql.cj.jdbc.Driver" -e DATABASE_PARAMS="characterEncoding=utf8&useSSL=false" \
-e DATABASE_HOST="192.168.1.125" -e DATABASE_PORT="3306" -e DATABASE_DATABASE="ds-k8s" \
-e DATABASE_USERNAME="root" -e DATABASE_PASSWORD="homch" \
-e ZOOKEEPER_QUORUM="192.168.1.123:8181" \
-e -e REGISTRY_SERVERS="192.168.1.123:8181" \
-p 12345:12345 \
apache/dolphinscheduler:mysql-driver all相关文章：
DolphinScheduler在Kubernetes体系中的技术实战
快速试用 Kubernetes 部署
docker--导出镜像 save/export、导入镜像 load/import区别及使用场景
DolphinScheduler部署中的一些坑（Docker & Kubernetes）
在 k8s 环境中使用 mysql 部署 dolphinscheduler (非 helm 的方式)
为者常成，行者常至
自由转载-非商用-非衍生-保持署名（创意共享3.0许可证）
