更新时间：2024-11-07 11:14:52
1 适用场景
该文档适用于如下场景进行产品扩容：
- 场景一：单节点扩容到三节点
- 场景二：单节点扩容到两节点
- 场景三：两节点扩容到三节点
- 场景四：三节点扩容到三节点以上
2 扩容说明
平台资源扩容主要指Kubernetes集群及Proton组件扩容，涉及如下内容：
• Kubernetes
• proton mariadb
• proton mongodb
• proton redis
• kafka
• zookeeper
• proton mq nsq
• proton policy engine
• proton-etcd
• nebula
说明：
- opensearch无需进行扩容
- 上述需扩容资源除proton-etcd外均通过proton-cli进行扩容
- kafka通过proton-cli进行副本扩容后，还需要修改数据副本
- nebula暂不支持扩容
- 上述组件均为proton内置组件，外置组件若需扩容请额外处理
文档中的扩容操作主要针对如下部署场景，若存在其他特殊场景请先与研发团队进行确认再进行扩容
- 单节点单master集群，所有服务均为单副本
- 两节点单master集群，所有服务均为单副本
- 三节点单master集群，所有服务均为单副本
- 三节点三master集群，所有服务均为三副本（全新部署三节点环境时默认设置为非三副本的服务除外）
集群master个数建议设置为奇数，最大推荐设置为3master。
多站点场景的扩容需要分别针对总站点和分站点进行扩容，扩容方式与单站点扩容相同。
2.1 其他说明
- 若对象存储（eceph）为独立部署，请参考对象存储独立部署与升级指导手册进行扩容
- 单节点扩容到两节点主要扩容对象为eceph
- 若未安装可观测性服务，则可跳过修改kafka数据副本（章节3.1.3）
3 准备工作
注意：扩容的节点（即新购机器）需要提前进行如下准备工作
3.1 安装操作系统
3.2 配置网络
3.2.1 配置节点ip
注意：节点ip需要与当前集群处于同一网段
- 打开配置文件：
vi /etc/sysconfig/network-scripts/ifcfg-{网卡名称}（网卡名称以实际为准） - 修改
BOOTPROTO = static；ONBOOT = yes；配置IPADDR，NETMASK，GATEWAY，请根据实际环境填写网络信息，下图仅为示例：
- 保存退出
- 重启网络：
systemctl restart network
3.2.2 配置内部ip
注意：
- 若当前集群未使用内部ip，可跳过该章节
- 配置的内部ip与已有集群节点内部ip网段需保持一致
- 内部ip请不要配置D类地址（224.0.0.0 ~ 239.255.255.255）
- 在/etc/sysconfig/network-scripts/ 下新增配置文件ifcfg-{网卡名称}:ip（网卡名称以实际为准），新增配置文件中添加DEVICE、PREFIX、BOOTPROTO、IPADDR、ONBOOT信息，网卡、ip、前缀请根据实际环境填写，示例如下：
- 保存退出
- 重启网络，执行
systemctl restart network或重启系统
3.3 设置主机名
执行如下命令，设置hostname：hostnamectl set-hostname {主机名称}
3.4 设置免密
注意：需要设置当前master免密登录到所有待扩容节点
在当前master节点依次执行命令设置免密登录到扩容节点：ssh-copy-id {ssh_id}，输入密码
3.5 安装依赖
注意：需要安装与当前集群节点版本一致的依赖，若当前节点上的依赖包已被删除，重新下载对应版本依赖包进行解压即可
- 在当前master节点，进入依赖包解压目录
- 执行命令安装依赖：
./install_deps.sh --hosts {ssh_ip1},{ssh_ip2}（若需要融合部署eceph，安装依赖命令为：./install_deps.sh --hosts {ssh_ip1},{ssh_ip2} --with-eceph-remote {ssh_ip1},{ssh_ip2}）
注意：ssh_ip1、ssh_ip2为待扩容节点
3.6 导入本地镜像（可选）
注意：若生产环境新部署时为proton 2.7.0（对应：AnyShare 7.0.4.8、AnyRobot 5.0.1.7、AnyDATA 2.0.1.8）及以后版本，则可跳过该章节；若环境为之前版本升级上来的，则需要按照如下步骤提前导入本地镜像（AnyFabric1.0.0.0对应proton 2.8.0，因此扩容时无需执行该章节）
第1步 在当前集群已有节点上执行：scp /usr/share/proton-cs/images.tar root@{待加入的新节点IP}:/root
第2步 在待加入的新节点上执行：docker load -i /root/images.tar，导入本地镜像
注意：上述步骤需要依次在每个待加入集群的新节点上执行
3.7 其他检查（可选）
请确认当前环境是否运行过补丁，并检查补丁镜像是否推送到registry仓库
docker images | grep {镜像名称}- 若存在两个同名镜像且镜像id不同，则需要将当前服务正在使用的镜像推送到registry仓库
- 执行命令推送镜像：
docker push registry.aishu.cn/{镜像路径}:{镜像tag}
注意：所有扩容操作均在当前集群master节点执行，proton-cli命令需要在依赖包解压目录下执行
4 单节点扩容到三节点
该场景下，原有环境集群节点为单节点，所有服务为单副本
- 获取配置，执行：
proton-cli get conf > cluster.yaml - 编辑配置文件，编辑后保存（配置文件示例如下）
- 进行平台资源扩容，执行：
proton-cli apply -f cluster.yaml - 等待执行完成，确认扩容结果
# 注：
# 1、该配置文件仅为示例说明，最终以实际环境中获取到的配置为准，配置文件中不涉及修改的组件无需关注
# 2、配置文件中节点ip、内部ip（若未使用内部ip，无需加上internal_ip这行配置）、hostmname信息请根据实际环境填写
# 3、单节点单master环境中数据服务均为单副本，扩容到三节点三master后，服务副本数与全新部署的三节点环境保持一致（nebula除外）
# 4、各服务hosts列表中的节点名称必须包含在nodes列表内
# 5、opensearch不扩容，proton-etcd之后单独执行扩容
# 6、若获取到的配置文件中不存在grafana和prometheus，则忽略这两个组件的修改
apiVersion: v1
nodes:
- internal_ip: 11.0.0.4
ip4: 10.3.91.101
name: node-91-101
- internal_ip: 11.0.0.29 # 新增节点，填写内部ip、节点ip、hostname信息
ip4: 10.3.91.102
name: node-91-102
- internal_ip: 11.0.0.30 # 新增节点，填写内部ip、节点ip、hostname信息
ip4: 10.3.91.103
name: node-91-103
cms: {}
cr:
local:
ha_ports:
chartmuseum: 15001
cr_manager: 15002
registry: 15000
rpm: 15003
hosts:
- node-91-101
- node-91-102 # 新增node-91-102将容器仓库扩容到两节点
ports:
chartmuseum: 5001
cr_manager: 5002
registry: 5000
rpm: 5003
storage: /sysvol/proton_data/cr_data
cs:
addons:
- node-exporter
- kube-state-metrics
cs_controller_dir: ./service-package
docker_data_dir: /sysvol/proton_data/cs_docker_data
etcd_data_dir: /sysvol/proton_data/cs_etcd_data
ha_port: 16643
host_network:
bip: 172.33.0.1/16
pod_network_cidr: 192.169.0.0/16
service_cidr: 10.96.0.0/12
ipFamilies:
- IPv4
master:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将kubernetes扩容到3 master
- node-91-103
provisioner: local
installer_service: {}
grafana:
data_path: /sysvol/grafana
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
kafka:
data_path: /sysvol/kafka/kafka_data
env: null
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将kafka扩容到3副本
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
nebula:
data_path: /sysvol/nebula
graphd: {}
hosts:
- node-91-101
metad: {}
password: ffa3b5b81a1344ac202f078e
storaged: {}
opensearch:
config:
hanlpRemoteextDict: http://ecoconfig-private.anyshare:32128/api/ecoconfig/v1/word-list/remote_ext_dict
jvmOptions: -Xmx8g -Xms8g
data_path: /anyshare/opensearch
hosts:
- node-91-101
mode: master
resources:
limits:
cpu: "8"
memory: 40Gi
requests:
cpu: "1"
memory: 2Gi
settings:
action.auto_create_index: -company*,+*
bootstrap.memory_lock: false
cluster.routing.allocation.disk.watermark.flood_stage: 70%
cluster.routing.allocation.disk.watermark.high: 65%
cluster.routing.allocation.disk.watermark.low: 60%
prometheus:
data_path: /sysvol/prometheus
hosts:
- node-91-101
- node-91-102 # 增加node-91-102将prometheus扩容到2副本
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
proton_etcd:
data_path: /sysvol/proton-etcd/proton-etcd_data
hosts:
- node-91-101
proton_mariadb:
admin_passwd: eisoo.com123
admin_user: root
config:
innodb_buffer_pool_size: 8G
resource_limits_memory: 24G
resource_requests_memory: 24G
data_path: /sysvol/mariadb
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mariadb扩容到3副本
- node-91-103
proton_mongodb:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/mongodb/mongodb_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mongodb扩容到3副本
- node-91-103
proton_mq_nsq:
data_path: /sysvol/mq-nsq/mq-nsq_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mq-nsq扩容到3副本
- node-91-103
proton_policy_engine:
data_path: /sysvol/policy-engine/policy-engine_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-policy-engine扩容到3副本
- node-91-103
proton_redis:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/redis/redis_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-redis扩容到3副本
- node-91-103
resource_connect_info:
etcd:
hosts: proton-etcd.resource
port: 2379
secret: etcdssl-secret
source_type: internal
mongodb:
auth_source: anyshare
hosts: mongodb-mongodb-0.mongodb-mongodb.resource,mongodb-mongodb-1.mongodb-mongodb.resource,mongodb-mongodb-2.mongodb-mongodb.resource
# mongodb.hosts更新为多副本连接
password: eisoo.com123
port: 28000
replica_set: rs0
source_type: internal
username: anyshare
mq:
mq_hosts: proton-mq-nsq-nsqd.resource
mq_lookupd_hosts: proton-mq-nsq-nsqlookupd.resource
mq_lookupd_port: 4161
mq_port: 4151
mq_type: nsq
source_type: internal
opensearch:
hosts: opensearch-master.resource
password: eisoo.com123
port: 9200
protocol: http
source_type: internal
username: admin
version: 7.10.0
policy_engine:
hosts: proton-policy-engine-proton-policy-engine-cluster.resource
port: 9800
source_type: internal
rds:
hosts: mariadb-mariadb-master.resource
hosts_read: mariadb-mariadb-cluster.resource
password: eisoo.com123
port: 3330
port_read: 3330
rds_type: MariaDB
source_type: internal
username: anyshare
redis:
connect_type: sentinel
master_group_name: mymaster
password: eisoo.com123
sentinel_hosts: proton-redis-proton-redis-sentinel.resource
sentinel_password: eisoo.com123
sentinel_port: 26379
sentinel_username: root
source_type: internal
username: root
zookeeper:
data_path: /sysvol/zookeeper/zookeeper_data
env: null
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将zookeeper扩容到3副本
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
docker_data_dir: /sysvol/proton_data/cs_docker_data
etcd_data_dir: /sysvol/proton_data/cs_etcd_data
ha_port: 16643
host_network:
bip: 172.33.0.1/16
pod_network_cidr: 192.169.0.0/16
service_cidr: 10.96.0.0/12
ipFamilies:
- IPv4
master:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将kubernetes扩容到3 master
- node-91-103
provisioner: local
installer_service: {}
grafana:
data_path: /sysvol/grafana
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
kafka:
data_path: /sysvol/kafka/kafka_data
env: null
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将kafka扩容到3副本
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
nebula:
data_path: /sysvol/nebula
graphd: {}
hosts:
- node-91-101
metad: {}
password: ffa3b5b81a1344ac202f078e
storaged: {}
opensearch:
config:
hanlpRemoteextDict: http://ecoconfig-private.anyshare:32128/api/ecoconfig/v1/word-list/remote_ext_dict
jvmOptions: -Xmx8g -Xms8g
data_path: /anyshare/opensearch
hosts:
- node-91-101
mode: master
resources:
limits:
cpu: "8"
memory: 40Gi
requests:
cpu: "1"
memory: 2Gi
settings:
action.auto_create_index: -company*,+*
bootstrap.memory_lock: false
cluster.routing.allocation.disk.watermark.flood_stage: 70%
cluster.routing.allocation.disk.watermark.high: 65%
cluster.routing.allocation.disk.watermark.low: 60%
prometheus:
data_path: /sysvol/prometheus
hosts:
- node-91-101
- node-91-102 # 增加node-91-102将prometheus扩容到2副本
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
proton_etcd:
data_path: /sysvol/proton-etcd/proton-etcd_data
hosts:
- node-91-101
proton_mariadb:
admin_passwd: eisoo.com123
admin_user: root
config:
innodb_buffer_pool_size: 8G
resource_limits_memory: 24G
resource_requests_memory: 24G
data_path: /sysvol/mariadb
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mariadb扩容到3副本
- node-91-103
proton_mongodb:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/mongodb/mongodb_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mongodb扩容到3副本
- node-91-103
proton_mq_nsq:
data_path: /sysvol/mq-nsq/mq-nsq_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mq-nsq扩容到3副本
- node-91-103
proton_policy_engine:
data_path: /sysvol/policy-engine/policy-engine_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-policy-engine扩容到3副本
- node-91-103
proton_redis:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/redis/redis_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-redis扩容到3副本
- node-91-103
resource_connect_info:
etcd:
hosts: proton-etcd.resource
port: 2379
secret: etcdssl-secret
source_type: internal
mongodb:
auth_source: anyshare
hosts: mongodb-mongodb-0.mongodb-mongodb.resource,mongodb-mongodb-1.mongodb-mongodb.resource,mongodb-mongodb-2.mongodb-mongodb.resource
# mongodb.hosts更新为多副本连接
password: eisoo.com123
port: 28000
replica_set: rs0
source_type: internal
username: anyshare
mq:
mq_hosts: proton-mq-nsq-nsqd.resource
mq_lookupd_hosts: proton-mq-nsq-nsqlookupd.resource
mq_lookupd_port: 4161
mq_port: 4151
mq_type: nsq
source_type: internal
opensearch:
hosts: opensearch-master.resource
password: eisoo.com123
port: 9200
protocol: http
source_type: internal
username: admin
version: 7.10.0
policy_engine:
hosts: proton-policy-engine-proton-policy-engine-cluster.resource
port: 9800
source_type: internal
rds:
hosts: mariadb-mariadb-master.resource
hosts_read: mariadb-mariadb-cluster.resource
password: eisoo.com123
port: 3330
port_read: 3330
rds_type: MariaDB
source_type: internal
username: anyshare
redis:
connect_type: sentinel
master_group_name: mymaster
password: eisoo.com123
sentinel_hosts: proton-redis-proton-redis-sentinel.resource
sentinel_password: eisoo.com123
sentinel_port: 26379
sentinel_username: root
source_type: internal
username: root
zookeeper:
data_path: /sysvol/zookeeper/zookeeper_data
env: null
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将zookeeper扩容到3副本
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
扩容结果确认
1.查看集群状态，三节点三master集群：kubectl get nodes
2.查看数据服务副本数及状态：kubectl get pod -n resource
proton-mariadb、proton-mongodb、proton-policy-engine、proton-mq-nsq、proton-redis、kafka、zookeeper均为3副本，且为running状态，（上述服务在进行数据同步过程中READY会显示为0/1，STATUS显示为Running，此时可继续进行后续操作，等待数据同步完成后READY会显示为1/1；处于Completed状态，READY显示为0/1的任务为正常现象）
- 查看集群配置：
proton-cli get conf
配置文件中节点为三节点，proton-mariadb、proton-mongodb、proton-policy-engine、proton-mq-nsq、proton-redis、kafka、zookeeper均显示为3副本
5 单节点扩容到两节点
该场景下，扩容主要针对对象存储（eceph）进行扩容，仅需要增加一个kubernetes的worker节点即可
- 获取配置，执行：
proton-cli get conf > cluster.yaml - 编辑配置文件，编辑后保存（配置文件示例如下）
- 进行平台资源扩容，执行：
proton-cli apply -f cluster.yaml - 等待执行完成，确认扩容结果
注：
# 1、该配置文件仅为示例说明，最终以实际环境中获取到的配置为准，配置文件中不涉及修改的组件无需关注
# 2、配置文件中节点ip、内部ip（若未使用内部ip，无需加上internal_ip这行配置）、hostmname信息请根据实际环境填写
# 3、单节点单master环境中数据服务均为单副本，扩容到两节点单master时，仅新增一个kubernetes的worker节点，其他服务不变
apiVersion: v1
nodes:
- internal_ip: 11.0.0.4
ip4: 10.3.91.101
name: node-91-101
- internal_ip: 11.0.0.29 # 新增节点，填写内部ip、节点ip、hostname信息
ip4: 10.3.91.102
name: node-91-102
cms: {}
cr:
local:
ha_ports:
chartmuseum: 15001
cr_manager: 15002
registry: 15000
rpm: 15003
hosts:
- node-91-101
ports:
chartmuseum: 5001
cr_manager: 5002
registry: 5000
rpm: 5003
storage: /sysvol/proton_data/cr_data
cs:
addons:
- node-exporter
- kube-state-metrics
cs_controller_dir: ./service-package
docker_data_dir: /sysvol/proton_data/cs_docker_data
etcd_data_dir: /sysvol/proton_data/cs_etcd_data
ha_port: 16643
host_network:
bip: 172.33.0.1/16
pod_network_cidr: 192.169.0.0/16
service_cidr: 10.96.0.0/12
ipFamilies:
- IPv4
master:
- node-91-101
provisioner: local
installer_service: {}
grafana:
data_path: /sysvol/grafana
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
kafka:
data_path: /sysvol/kafka/kafka_data
env: null
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
nebula:
data_path: /sysvol/nebula
graphd: {}
hosts:
- node-91-101
metad: {}
password: ffa3b5b81a1344ac202f078e
storaged: {}
opensearch:
config:
hanlpRemoteextDict: http://ecoconfig-private.anyshare:32128/api/ecoconfig/v1/word-list/remote_ext_dict
jvmOptions: -Xmx8g -Xms8g
data_path: /anyshare/opensearch
hosts:
- node-91-101
mode: master
resources:
limits:
cpu: "8"
memory: 40Gi
requests:
cpu: "1"
memory: 2Gi
settings:
action.auto_create_index: -company*,+*
bootstrap.memory_lock: false
cluster.routing.allocation.disk.watermark.flood_stage: 70%
cluster.routing.allocation.disk.watermark.high: 65%
cluster.routing.allocation.disk.watermark.low: 60%
prometheus:
data_path: /sysvol/prometheus
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
proton_etcd:
data_path: /sysvol/proton-etcd/proton-etcd_data
hosts:
- node-91-101
proton_mariadb:
admin_passwd: eisoo.com123
admin_user: root
config:
innodb_buffer_pool_size: 8G
resource_limits_memory: 24G
resource_requests_memory: 24G
data_path: /sysvol/mariadb
hosts:
- node-91-101
proton_mongodb:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/mongodb/mongodb_data
hosts:
- node-91-101
proton_mq_nsq:
data_path: /sysvol/mq-nsq/mq-nsq_data
hosts:
- node-91-101
proton_policy_engine:
data_path: /sysvol/policy-engine/policy-engine_data
hosts:
- node-91-101
proton_redis:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/redis/redis_data
hosts:
- node-91-101
resource_connect_info:
etcd:
hosts: proton-etcd.resource
port: 2379
secret: etcdssl-secret
source_type: internal
mongodb:
auth_source: anyshare
hosts: mongodb-mongodb-0.mongodb-mongodb.resource
password: eisoo.com123
port: 28000
replica_set: rs0
source_type: internal
username: anyshare
mq:
mq_hosts: proton-mq-nsq-nsqd.resource
mq_lookupd_hosts: proton-mq-nsq-nsqlookupd.resource
mq_lookupd_port: 4161
mq_port: 4151
mq_type: nsq
source_type: internal
opensearch:
hosts: opensearch-master.resource
password: eisoo.com123
port: 9200
protocol: http
source_type: internal
username: admin
version: 7.10.0
policy_engine:
hosts: proton-policy-engine-proton-policy-engine-cluster.resource
port: 9800
source_type: internal
rds:
hosts: mariadb-mariadb-master.resource
hosts_read: mariadb-mariadb-cluster.resource
password: eisoo.com123
port: 3330
port_read: 3330
rds_type: MariaDB
source_type: internal
username: anyshare
redis:
connect_type: sentinel
master_group_name: mymaster
password: eisoo.com123
sentinel_hosts: proton-redis-proton-redis-sentinel.resource
sentinel_password: eisoo.com123
sentinel_port: 26379
sentinel_username: root
source_type: internal
username: root
zookeeper:
data_path: /sysvol/zookeeper/zookeeper_data
env: null
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
扩容结果确认
查看集群状态，两节点单master集群：kubectl get nodes
6 两节点扩容到三节
该场景下，原有环境集群节点为两节点，master节点为1个，所有服务为单副本
- 获取配置，执行：
proton-cli get conf > cluster.yaml - 编辑配置文件，编辑后保存（配置文件示例如下）
- 进行平台资源扩容，执行：
proton-cli apply -f cluster.yaml - 等待执行完成，确认扩容结果
注：
# 1、该配置文件仅为示例说明，最终以实际环境中获取到的配置为准，配置文件中不涉及修改的组件无需关注
# 2、配置文件中节点ip、内部ip（若未使用内部ip，无需加上internal_ip这行配置）、hostmname信息请根据实际环境填写
# 3、两节点单master环境中数据服务均为单副本，扩容到三节点三master后，服务副本数与全新部署的三节点环境保持一致（nebula除外）
# 4、各服务hosts列表中的节点名称必须包含在nodes列表内
# 5、opensearch不扩容，proton-etcd之后单独执行扩容
# 6、若获取到的配置文件中不存在grafana和prometheus，则忽略这两个组件的修改
apiVersion: v1
nodes:
- internal_ip: 11.0.0.4
ip4: 10.3.91.101
name: node-91-101
- internal_ip: 11.0.0.29
ip4: 10.3.91.102
name: node-91-102
- internal_ip: 11.0.0.30 # 新增节点，填写内部ip、节点ip、hostname信息
ip4: 10.3.91.103
name: node-91-103
cms: {}
cr:
local:
ha_ports:
chartmuseum: 15001
cr_manager: 15002
registry: 15000
rpm: 15003
hosts:
- node-91-101
- node-91-102 # 新增node-91-102将容器仓库扩容到两节点
ports:
chartmuseum: 5001
cr_manager: 5002
registry: 5000
rpm: 5003
storage: /sysvol/proton_data/cr_data
cs:
addons:
- node-exporter
- kube-state-metrics
cs_controller_dir: ./service-package
docker_data_dir: /sysvol/proton_data/cs_docker_data
etcd_data_dir: /sysvol/proton_data/cs_etcd_data
ha_port: 16643
host_network:
bip: 172.33.0.1/16
pod_network_cidr: 192.169.0.0/16
service_cidr: 10.96.0.0/12
ipFamilies:
- IPv4
master:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将kubernetes扩容到3 master
- node-91-103
provisioner: local
installer_service: {}
grafana:
data_path: /sysvol/grafana
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
kafka:
data_path: /sysvol/kafka/kafka_data
env: null
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将kafka扩容到3副本
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
nebula:
data_path: /sysvol/nebula
graphd: {}
hosts:
- node-91-101
metad: {}
password: ffa3b5b81a1344ac202f078e
storaged: {}
opensearch:
config:
hanlpRemoteextDict: http://ecoconfig-private.anyshare:32128/api/ecoconfig/v1/word-list/remote_ext_dict
jvmOptions: -Xmx8g -Xms8g
data_path: /anyshare/opensearch
hosts:
- node-91-101
mode: master
resources:
limits:
cpu: "8"
memory: 40Gi
requests:
cpu: "1"
memory: 2Gi
settings:
action.auto_create_index: -company*,+*
bootstrap.memory_lock: false
cluster.routing.allocation.disk.watermark.flood_stage: 70%
cluster.routing.allocation.disk.watermark.high: 65%
cluster.routing.allocation.disk.watermark.low: 60%
prometheus:
data_path: /sysvol/prometheus
hosts:
- node-91-101
- node-91-102 # 增加node-91-102将prometheus扩容到2副本
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
proton_etcd:
data_path: /sysvol/proton-etcd/proton-etcd_data
hosts:
- node-91-101
proton_mariadb:
admin_passwd: eisoo.com123
admin_user: root
config:
innodb_buffer_pool_size: 8G
resource_limits_memory: 24G
resource_requests_memory: 24G
data_path: /sysvol/mariadb
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mariadb扩容到3副本
- node-91-103
proton_mongodb:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/mongodb/mongodb_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mongodb扩容到3副本
- node-91-103
proton_mq_nsq:
data_path: /sysvol/mq-nsq/mq-nsq_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mq-nsq扩容到3副本
- node-91-103
proton_policy_engine:
data_path: /sysvol/policy-engine/policy-engine_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-policy-engine扩容到3副本
- node-91-103
proton_redis:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/redis/redis_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-redis扩容到3副本
- node-91-103
resource_connect_info:
etcd:
hosts: proton-etcd.resource
port: 2379
secret: etcdssl-secret
source_type: internal
mongodb:
auth_source: anyshare
hosts: mongodb-mongodb-0.mongodb-mongodb.resource,mongodb-mongodb-1.mongodb-mongodb.resource,mongodb-mongodb-2.mongodb-mongodb.resource
# mongodb.hosts更新为多副本连接
password: eisoo.com123
port: 28000
replica_set: rs0
source_type: internal
username: anyshare
mq:
mq_hosts: proton-mq-nsq-nsqd.resource
mq_lookupd_hosts: proton-mq-nsq-nsqlookupd.resource
mq_lookupd_port: 4161
mq_port: 4151
mq_type: nsq
source_type: internal
opensearch:
hosts: opensearch-master.resource
password: eisoo.com123
port: 9200
protocol: http
source_type: internal
username: admin
version: 7.10.0
policy_engine:
hosts: proton-policy-engine-proton-policy-engine-cluster.resource
port: 9800
source_type: internal
rds:
hosts: mariadb-mariadb-master.resource
hosts_read: mariadb-mariadb-cluster.resource
password: eisoo.com123
port: 3330
port_read: 3330
rds_type: MariaDB
source_type: internal
username: anyshare
redis:
connect_type: sentinel
master_group_name: mymaster
password: eisoo.com123
sentinel_hosts: proton-redis-proton-redis-sentinel.resource
sentinel_password: eisoo.com123
sentinel_port: 26379
sentinel_username: root
source_type: internal
username: root
zookeeper:
data_path: /sysvol/zookeeper/zookeeper_data
env: null
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将zookeeper扩容到3副本
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
扩容结果确认
- 查看集群状态，三节点三master集群：kubectl get nodes
- 查看数据服务副本数及状态：kubectl get pod -n resource
proton-mariadb、proton-mongodb、proton-policy-engine、proton-mq-nsq、proton-redis、kafka、zookeeper均为三副本，且为running状态，（上述服务在进行数据同步过程中READY会显示为0/1，STATUS显示为Running，此时可继续进行后续操作，等待数据同步完成后READY会显示为1/1；处于Completed状态，READY显示为0/1的任务为正常现象）
- 查看集群配置：proton-cli get conf
配置文件中节点为三节点，proton-mariadb、proton-mongodb、proton-policy-engine、proton-mq-nsq、proton-redis、kafka、zookeeper均显示为三副本
7 三节点单master扩容到三节点三master
该场景下，原有环境集群节点为三节点，master节点为1个，所有服务为单副本
- 获取配置，执行：
proton-cli get conf > cluster.yaml - 编辑配置文件，编辑后保存（配置文件示例如下）
- 进行平台资源扩容，执行：
proton-cli apply -f cluster.yaml - 等待执行完成，确认扩容结果
# 注：
# 1、该配置文件仅为示例说明，最终以实际环境中获取到的配置为准，配置文件中不涉及修改的组件无需关注
# 2、配置文件中节点ip、内部ip（若未使用内部ip，无需加上internal_ip这行配置）、hostmname信息请根据实际环境填写
# 3、三节点单master环境中数据服务均为单副本，扩容到三节点三master后，服务副本数与全新部署的三节点环境保持一致（nebula除外）
# 4、各服务hosts列表中的节点名称必须包含在nodes列表内
# 5、opensearch不扩容，proton-etcd之后单独执行扩容
# 6、若获取到的配置文件中不存在grafana和prometheus，则忽略这两个组件的修改
apiVersion: v1
nodes:
- internal_ip: 11.0.0.4
ip4: 10.3.91.101
name: node-91-101
- internal_ip: 11.0.0.29
ip4: 10.3.91.102
name: node-91-102
- internal_ip: 11.0.0.30
ip4: 10.3.91.103
name: node-91-103
cms: {}
cr:
local:
ha_ports:
chartmuseum: 15001
cr_manager: 15002
registry: 15000
rpm: 15003
hosts:
- node-91-101
- node-91-102 # 新增node-91-102将容器仓库扩容到两节点
ports:
chartmuseum: 5001
cr_manager: 5002
registry: 5000
rpm: 5003
storage: /sysvol/proton_data/cr_data
cs:
addons:
- node-exporter
- kube-state-metrics
cs_controller_dir: ./service-package
docker_data_dir: /sysvol/proton_data/cs_docker_data
etcd_data_dir: /sysvol/proton_data/cs_etcd_data
ha_port: 16643
host_network:
bip: 172.33.0.1/16
pod_network_cidr: 192.169.0.0/16
service_cidr: 10.96.0.0/12
ipFamilies:
- IPv4
master:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将kubernetes扩容到3 master
- node-91-103
provisioner: local
installer_service: {}
grafana:
data_path: /sysvol/grafana
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
kafka:
data_path: /sysvol/kafka/kafka_data
env: null
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将kafka扩容到3副本
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
nebula:
data_path: /sysvol/nebula
graphd: {}
hosts:
- node-91-101
metad: {}
password: ffa3b5b81a1344ac202f078e
storaged: {}
opensearch:
config:
hanlpRemoteextDict: http://ecoconfig-private.anyshare:32128/api/ecoconfig/v1/word-list/remote_ext_dict
jvmOptions: -Xmx8g -Xms8g
data_path: /anyshare/opensearch
hosts:
- node-91-101
mode: master
resources:
limits:
cpu: "8"
memory: 40Gi
requests:
cpu: "1"
memory: 2Gi
settings:
action.auto_create_index: -company*,+*
bootstrap.memory_lock: false
cluster.routing.allocation.disk.watermark.flood_stage: 70%
cluster.routing.allocation.disk.watermark.high: 65%
cluster.routing.allocation.disk.watermark.low: 60%
prometheus:
data_path: /sysvol/prometheus
hosts:
- node-91-101
- node-91-102 # 增加node-91-102将prometheus扩容到2副本
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
proton_etcd:
data_path: /sysvol/proton-etcd/proton-etcd_data
hosts:
- node-91-101
proton_mariadb:
admin_passwd: eisoo.com123
admin_user: root
config:
innodb_buffer_pool_size: 8G
resource_limits_memory: 24G
resource_requests_memory: 24G
data_path: /sysvol/mariadb
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mariadb扩容到3副本
- node-91-103
proton_mongodb:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/mongodb/mongodb_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mongodb扩容到3副本
- node-91-103
proton_mq_nsq:
data_path: /sysvol/mq-nsq/mq-nsq_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-mq-nsq扩容到3副本
- node-91-103
proton_policy_engine:
data_path: /sysvol/policy-engine/policy-engine_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-policy-engine扩容到3副本
- node-91-103
proton_redis:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/redis/redis_data
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将proton-redis扩容到3副本
- node-91-103
resource_connect_info:
etcd:
hosts: proton-etcd.resource
port: 2379
secret: etcdssl-secret
source_type: internal
mongodb:
auth_source: anyshare
hosts: mongodb-mongodb-0.mongodb-mongodb.resource,mongodb-mongodb-1.mongodb-mongodb.resource,mongodb-mongodb-2.mongodb-mongodb.resource
# mongodb.hosts更新为多副本连接
password: eisoo.com123
port: 28000
replica_set: rs0
source_type: internal
username: anyshare
mq:
mq_hosts: proton-mq-nsq-nsqd.resource
mq_lookupd_hosts: proton-mq-nsq-nsqlookupd.resource
mq_lookupd_port: 4161
mq_port: 4151
mq_type: nsq
source_type: internal
opensearch:
hosts: opensearch-master.resource
password: eisoo.com123
port: 9200
protocol: http
source_type: internal
username: admin
version: 7.10.0
policy_engine:
hosts: proton-policy-engine-proton-policy-engine-cluster.resource
port: 9800
source_type: internal
rds:
hosts: mariadb-mariadb-master.resource
hosts_read: mariadb-mariadb-cluster.resource
password: eisoo.com123
port: 3330
port_read: 3330
rds_type: MariaDB
source_type: internal
username: anyshare
redis:
connect_type: sentinel
master_group_name: mymaster
password: eisoo.com123
sentinel_hosts: proton-redis-proton-redis-sentinel.resource
sentinel_password: eisoo.com123
sentinel_port: 26379
sentinel_username: root
source_type: internal
username: root
zookeeper:
data_path: /sysvol/zookeeper/zookeeper_data
env: null
hosts:
- node-91-101
- node-91-102 # 增加node-91-102及node-91-103将zookeeper扩容到3副本
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
扩容结果确认
- 查看集群状态，三节点三master集群：kubectl get nodes
- 查看数据服务副本数及状态：kubectl get pod -n resource
proton-mariadb、proton-mongodb、proton-policy-engine、proton-mq-nsq、proton-redis、kafka、zookeeper均为三副本，且为running状态，（上述服务在进行数据同步过程中READY会显示为0/1，STATUS显示为Running，此时可继续进行后续操作，等待数据同步完成后READY会显示为1/1；处于Completed状态，READY显示为0/1的任务为正常现象）
- 查看集群配置：proton-cli get conf
配置文件中节点为三节点，proton-mariadb、proton-mongodb、proton-policy-engine、proton-mq-nsq、proton-redis、kafka、zookeeper均显示为三副本
8 三节点三master扩容到三节点以上
该场景下，原有环境集群节点为三节点，master节点为3个，cr服务为2个，opensearch为单副本，其他服务均为三副本，此时扩容，仅需要新增kubernetes的worker节点，master节点及数据服务无需再进行扩容
- 获取配置，执行：
proton-cli get conf > cluster.yaml - 编辑配置文件，编辑后保存（配置文件示例如下）
- 进行平台资源扩容，执行：
proton-cli apply -f cluster.yaml - 等待执行完成，确认扩容结果
# 注：
# 1、该配置文件仅为示例说明，最终以实际环境中获取到的配置为准，配置文件中不涉及修改的组件无需关注
# 2、配置文件中节点ip、内部ip（若未使用内部ip，无需加上internal_ip这行配置）、hostmname信息请根据实际环境填写
# 3、三节点三master扩容到三节点以上仅需扩容kubernetes的worker节点
apiVersion: v1
nodes:
- internal_ip: 11.0.0.4
ip4: 10.3.91.101
name: node-91-101
- internal_ip: 11.0.0.29
ip4: 10.3.91.102
name: node-91-102
- internal_ip: 11.0.0.30
ip4: 10.3.91.103
name: node-91-103
- internal_ip: 11.0.0.31 # 新增节点，填写内部ip、节点ip、hostname信息；若需增加多个节点，在nodes列表中依次添加即可
ip4: 10.3.91.104
name: node-91-104
cms: {}
cr:
local:
ha_ports:
chartmuseum: 15001
cr_manager: 15002
registry: 15000
rpm: 15003
hosts:
- node-91-101
- node-91-102
ports:
chartmuseum: 5001
cr_manager: 5002
registry: 5000
rpm: 5003
storage: /sysvol/proton_data/cr_data
cs:
addons:
- node-exporter
- kube-state-metrics
cs_controller_dir: ./service-package
docker_data_dir: /sysvol/proton_data/cs_docker_data
etcd_data_dir: /sysvol/proton_data/cs_etcd_data
ha_port: 16643
host_network:
bip: 172.33.0.1/16
pod_network_cidr: 192.169.0.0/16
service_cidr: 10.96.0.0/12
ipFamilies:
- IPv4
master:
- node-91-101
- node-91-102
- node-91-103
provisioner: local
installer_service: {}
grafana:
data_path: /sysvol/grafana
hosts:
- node-91-101
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
kafka:
data_path: /sysvol/kafka/kafka_data
env: null
hosts:
- node-91-101
- node-91-102
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
nebula:
data_path: /sysvol/nebula
graphd: {}
hosts:
- node-91-101
metad: {}
password: ffa3b5b81a1344ac202f078e
storaged: {}
opensearch:
config:
hanlpRemoteextDict: http://ecoconfig-private.anyshare:32128/api/ecoconfig/v1/word-list/remote_ext_dict
jvmOptions: -Xmx8g -Xms8g
data_path: /anyshare/opensearch
hosts:
- node-91-101
mode: master
resources:
limits:
cpu: "8"
memory: 40Gi
requests:
cpu: "1"
memory: 2Gi
settings:
action.auto_create_index: -company*,+*
bootstrap.memory_lock: false
cluster.routing.allocation.disk.watermark.flood_stage: 70%
cluster.routing.allocation.disk.watermark.high: 65%
cluster.routing.allocation.disk.watermark.low: 60%
prometheus:
data_path: /sysvol/prometheus
hosts:
- node-91-101
- node-91-102
resources:
limits:
cpu: "1"
memory: 3Gi
requests:
cpu: 500m
memory: 2Gi
proton_etcd:
data_path: /sysvol/proton-etcd/proton-etcd_data
hosts:
- node-91-101
- node-91-102
- node-91-103
proton_mariadb:
admin_passwd: eisoo.com123
admin_user: root
config:
innodb_buffer_pool_size: 8G
resource_limits_memory: 24G
resource_requests_memory: 24G
data_path: /sysvol/mariadb
hosts:
- node-91-101
- node-91-102
- node-91-103
proton_mongodb:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/mongodb/mongodb_data
hosts:
- node-91-101
- node-91-102
- node-91-103
proton_mq_nsq:
data_path: /sysvol/mq-nsq/mq-nsq_data
hosts:
- node-91-101
- node-91-102
- node-91-103
proton_policy_engine:
data_path: /sysvol/policy-engine/policy-engine_data
hosts:
- node-91-101
- node-91-102
- node-91-103
proton_redis:
admin_passwd: eisoo.com123
admin_user: root
data_path: /sysvol/redis/redis_data
hosts:
- node-91-101
- node-91-102
- node-91-103
resource_connect_info:
etcd:
hosts: proton-etcd.resource
port: 2379
secret: etcdssl-secret
source_type: internal
mongodb:
auth_source: anyshare
hosts: mongodb-mongodb-0.mongodb-mongodb.resource,mongodb-mongodb-1.mongodb-mongodb.resource,mongodb-mongodb-2.mongodb-mongodb.resource
password: eisoo.com123
port: 28000
replica_set: rs0
source_type: internal
username: anyshare
mq:
mq_hosts: proton-mq-nsq-nsqd.resource
mq_lookupd_hosts: proton-mq-nsq-nsqlookupd.resource
mq_lookupd_port: 4161
mq_port: 4151
mq_type: nsq
source_type: internal
opensearch:
hosts: opensearch-master.resource
password: eisoo.com123
port: 9200
protocol: http
source_type: internal
username: admin
version: 7.10.0
policy_engine:
hosts: proton-policy-engine-proton-policy-engine-cluster.resource
port: 9800
source_type: internal
rds:
hosts: mariadb-mariadb-master.resource
hosts_read: mariadb-mariadb-cluster.resource
password: eisoo.com123
port: 3330
port_read: 3330
rds_type: MariaDB
source_type: internal
username: anyshare
redis:
connect_type: sentinel
master_group_name: mymaster
password: eisoo.com123
sentinel_hosts: proton-redis-proton-redis-sentinel.resource
sentinel_password: eisoo.com123
sentinel_port: 26379
sentinel_username: root
source_type: internal
username: root
zookeeper:
data_path: /sysvol/zookeeper/zookeeper_data
env: null
hosts:
- node-91-101
- node-91-102
- node-91-103
resources:
limits:
cpu: "1"
memory: 2Gi
requests:
cpu: 500m
memory: 1Gi
扩容结果确认
查看集群状态：kubectl get nodes
9 Proton-etcd 扩容
该章节仅需要在单节点扩展到三节点、两节点扩展到三节点、三节点单master扩展到三节点三master的场景下执行
第1步 查看proton-etcd数据，执行：kubectl -n resource exec -it proton-etcd-0 -- etcdctl --cacert=/var/etcd/peer/ca.crt --cert=/var/etcd/peer/peer.crt --key=/var/etcd/peer/peer.key get / --prefix
第2步 获取proton-etcd配置，执行：helm3 get values proton-etcd -n resource > proton-etcd.yaml
第3步 编辑proton-etcd配置文件，修改副本数和节点信息
副本数修改为3，添加storage.local中1和2，host请根据环境实际情况填写，1和2中的path请与0保持一致，如图所示：
第4步 上传proton-etcd扩容工具到服务器（请根据服务器架构下载对应的脚本）脚本路径如下：
第5步 给脚本赋予执行权限：chmod +x ./proton-etcd-repair-aarch64-v20230925（请注意使用对应架构脚本）
第6步 执行proton-etcd扩容：./proton-etcd-repair-aarch64-v20230925 -values proton-etcd.yaml（请注意使用对应架构脚本），等待扩容完成
第7步 执行kubectl edit secret -n resource persist-component-manage-components,复制下面圈起来的内容
使用上面复制出来的内容执行下面截图命令
编辑文件vi component-manage.etcd.yaml，将被扩容的节点添加到该配置文件，注意：添加节点的顺序要和proton-cli get conf 获取到etcd节点顺序一致，如下截图：
文件编辑完成后执行下面截图里命令进行base64编码
复制上一步的base64编码，替换kubectl edit secret -n resource persist-component-manage-components命令获取的proton-etcd字段值，保存后退出
第7步 更新proton-policy-engine中proton-etcd的endpoints，依次执行：proton-cli get conf > cluster.yaml；proton-cli apply -f cluster.yaml
第9步 确认proton-etcd副本数及运行状态，执行：kubectl get pod -n resource | grep proton-etcd
proton-etcd为三副本，且为running状态，如图所示：
第10步 确认proton-etcd配置，执行：proton-cli get conf
配置文件中proton-etcd应显示为三副本，如图：
第11步 确认proton-etcd数据，执行：kubectl -n resource exec -it proton-etcd-1 -- etcdctl --cacert=/var/etcd/peer/ca.crt --cert=/var/etcd/peer/peer.crt --key=/var/etcd/peer/peer.key get / --prefix
获取到的数据与扩容前的数据应保持一致
第12步 确认proton-policy-engine中proton-etcd的endpoints，执行：helm3 get values proton-policy-engine -n resource，proton-policy-engine配置中etcd.endpoints应包含proton-etcd-0、proton-etcd-1和proton-etcd-2
10 修改kafka数据副本
该章节仅需要在单节点扩展到三节点、两节点扩展到三节点、三节点单master扩展到三节点三master的场景下执行，kafka的pod已被扩容为3副本
第1步 查看当前topic：kubectl exec -it -n resource kafka-0 -- /opt/kafka/bin/kafka-topics.sh --list --bootstrap-server localhost:9097 --command-config /opt/kafka/config/client.properties
第2步 查看已存在topic的分区数和副本数：kubectl exec -it kafka-0 -n resource -- /opt/kafka/bin/kafka-topics.sh --describe --bootstrap-server localhost:9097 --command-config /opt/kafka/config/client.properties --topic kafka_topic（以查看topic：kafka_topic的分区数和副本数为例），此时topic的分区数和副本数均为1
第3步 修改已存在topic的分区数和副本数，上传修改工具到服务器，修改脚本权限，执行修改（请使用对应架构脚本），脚本目录如下：
修改脚本权限：chmod +x updateKafkaTopicReplicas-amd64-v20230925
执行修改：./updateKafkaTopicReplicas-amd64-v20230925
第4步 修改后查看已存在topic的分区数和副本数，此时topic的分区数为3，副本数为2（以查看topic：kafka_topic的分区数和副本数为例）
11 部署工作台扩容
该章节仅需要在单节点扩展到三节点、两节点扩展到三节点、三节点单master扩展到三节点三master的场景下执行
注意：部署工作台更新时会同步更新依赖MongoDB的相关服务的MongoDB连接信息
第1步 更新部署工作台，登录部署工作台，进入【服务管理】--【服务部署】，选中DeploymentStudio，点击【更新】
第2步 选中当前版本，点击【下一步】，填写配置项，将服务副本数改为3，点击【SUBMIT】，点击【下一步】，点击【确定】，等待更新完成
第3步 确认部署工作台扩容结果，执行：kubectl get pod -n anyshare，服务为三副本且处于正常运行状态
12 配置高可用
若当前环境已配置过高可用，可跳过该章节
第1步 登录部署工作台
第2步 进入【环境与资源】-- 【服务器】，点击【设置高可用节点】-- 点击【添加】，选中节点，点击确定
第3步 填写访问ip，前缀及网卡，点击【确定】
第4步 更新访问配置，进入【环境与资源】-- 【访问配置】，点击【修改访问配置】，将访问地址修改为第3步设置好的访问ip（也可设置为域名，设置为域名时，需要配置好域名与访问ip的解析规则），点击【保存】
第5步 更新证书，进入【环境与资源】-- 【访问配置】，点击【配置证书】-- 点击【生成自签名证书】
< 上一篇：
下一篇： >