提交代码

This commit is contained in:
zlwu committed 2025-08-31 19:38:31 +08:00
1 parent 3134870b68
commit c20df56dcd
26 files changed
+2931 -37

No files matched your search

@@ -115,8 +115,8 @@ spring:
idleTimeout: 300000
# 此属性控制池中连接的最长生命周期,值0表示无限生命周期,默认30分钟(减少生命周期)
maxLifetime: 900000
# 多久检查一次连接的活性(增加间隔,减少检测频率)
keepaliveTime: 60000
# 多久检查一次连接的活性(增加间隔,减少检测频率,减少日志输出)
keepaliveTime: 300000
# 连接测试查询(验证连接有效性)
# connectionTestQuery: SELECT 11
@@ -0,0 +1,44 @@
# 生产环境配置 - 9900端口实例
server:
port: 9900
servlet:
context-path: /
# 实例标识
spring:
application:
name: RuoYi-Vue-Plus-9900
# 日志文件配置
logging:
file:
name: logs/ruoyi-admin-9900.log
pattern:
file: "%d{yyyy-MM-dd HH:mm:ss.SSS} [%thread] %-5level %logger{50} - %msg%n"
# 监控中心配置 - 端口漂移
spring.boot.admin.client:
enabled: true
url: http://172.16.0.67:9090/admin
instance:
service-host-type: IP
metadata:
username: ${spring.boot.admin.client.username}
userpassword: ${spring.boot.admin.client.password}
instance.port: 9900
instance.name: ruoyi-admin-9900
username: @monitor.username@
password: @monitor.password@
# SnailJob 配置 - 端口漂移
snail-job:
enabled: true
group: "ruoyi_group_9900"
token: "SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT"
server:
host: 172.16.0.67
port: 17888
namespace: "prod"
port: 29900
host:
rpc-type: grpc
@@ -0,0 +1,44 @@
# 生产环境配置 - 9901端口实例
server:
port: 9901
servlet:
context-path: /
# 实例标识
spring:
application:
name: RuoYi-Vue-Plus-9901
# 日志文件配置
logging:
file:
name: logs/ruoyi-admin-9901.log
pattern:
file: "%d{yyyy-MM-dd HH:mm:ss.SSS} [%thread] %-5level %logger{50} - %msg%n"
# 监控中心配置 - 端口漂移
spring.boot.admin.client:
enabled: true
url: http://172.16.0.67:9090/admin
instance:
service-host-type: IP
metadata:
username: ${spring.boot.admin.client.username}
userpassword: ${spring.boot.admin.client.password}
instance.port: 9901
instance.name: ruoyi-admin-9901
username: @monitor.username@
password: @monitor.password@
# SnailJob 配置 - 端口漂移
snail-job:
enabled: true
group: "ruoyi_group_9901"
token: "SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT"
server:
host: 172.16.0.67
port: 17888
namespace: "prod"
port: 29901
host:
rpc-type: grpc
@@ -0,0 +1,44 @@
# 生产环境配置 - 9902端口实例
server:
port: 9902
servlet:
context-path: /
# 实例标识
spring:
application:
name: RuoYi-Vue-Plus-9902
# 日志文件配置
logging:
file:
name: logs/ruoyi-admin-9902.log
pattern:
file: "%d{yyyy-MM-dd HH:mm:ss.SSS} [%thread] %-5level %logger{50} - %msg%n"
# 监控中心配置 - 端口漂移
spring.boot.admin.client:
enabled: true
url: http://172.16.0.67:9090/admin
instance:
service-host-type: IP
metadata:
username: ${spring.boot.admin.client.username}
userpassword: ${spring.boot.admin.client.password}
instance.port: 9902
instance.name: ruoyi-admin-9902
username: @monitor.username@
password: @monitor.password@
# SnailJob 配置 - 端口漂移
snail-job:
enabled: true
group: "ruoyi_group_9902"
token: "SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT"
server:
host: 172.16.0.67
port: 17888
namespace: "prod"
port: 29902
host:
rpc-type: grpc
@@ -0,0 +1,44 @@
# 生产环境配置 - 9903端口实例
server:
port: 9903
servlet:
context-path: /
# 实例标识
spring:
application:
name: RuoYi-Vue-Plus-9903
# 日志文件配置
logging:
file:
name: logs/ruoyi-admin-9903.log
pattern:
file: "%d{yyyy-MM-dd HH:mm:ss.SSS} [%thread] %-5level %logger{50} - %msg%n"
# 监控中心配置 - 端口漂移
spring.boot.admin.client:
enabled: true
url: http://172.16.0.67:9090/admin
instance:
service-host-type: IP
metadata:
username: ${spring.boot.admin.client.username}
userpassword: ${spring.boot.admin.client.password}
instance.port: 9903
instance.name: ruoyi-admin-9903
username: @monitor.username@
password: @monitor.password@
# SnailJob 配置 - 端口漂移
snail-job:
enabled: true
group: "ruoyi_group_9903"
token: "SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT"
server:
host: 172.16.0.67
port: 17888
namespace: "prod"
port: 29903
host:
rpc-type: grpc
@@ -0,0 +1,353 @@
--- # 临时文件存储位置 避免临时文件被系统清理报错
spring.servlet.multipart.location: /ruoyi/server/temp
--- # 监控中心配置
spring.boot.admin.client:
# 增加客户端开关
enabled: true
url: http://47.121.125.41:9090/admin
instance:
service-host-type: IP
metadata:
username: ${spring.boot.admin.client.username}
userpassword: ${spring.boot.admin.client.password}
username: @monitor.username@
password: @monitor.password@
--- # snail-job 配置
snail-job:
enabled: true
# 需要在 SnailJob 后台组管理创建对应名称的组,然后创建任务的时候选择对应的组,才能正确分派任务
group: "ruoyi_group"
# SnailJob 接入验证令牌 详见 script/sql/ry_job.sql `sj_group_config`表
token: "SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT"
server:
host: 47.121.125.41
port: 17888
# 命名空间UUID 详见 script/sql/ry_job.sql `sj_namespace`表`unique_id`字段
namespace: ${spring.profiles.active}
# 随主应用端口漂移
port: 2${server.port}
# 客户端ip指定
host:
# RPC类型: netty, grpc
rpc-type: grpc
--- # 数据源配置
spring:
datasource:
type: com.zaxxer.hikari.HikariDataSource
# 动态数据源文档 https://www.kancloud.cn/tracy5546/dynamic-datasource/content
dynamic:
# 性能分析插件(有性能损耗 不建议生产环境使用)
p6spy: false
# 设置默认的数据源或者数据源组,默认值即为 master
primary: master
# 严格模式 匹配不到数据源则报错
strict: true
datasource:
# 主库数据源
master:
type: ${spring.datasource.type}
driverClassName: com.mysql.cj.jdbc.Driver
# jdbc 所有参数配置参考 https://lionli.blog.csdn.net/article/details/122018562
# rewriteBatchedStatements=true 批处理优化 大幅提升批量插入更新删除性能(对数据库有性能损耗 使用批量操作应考虑性能问题)
url: jdbc:mysql://47.99.164.66:3306/hzdyjs_xg?useUnicode=true&characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=true&serverTimezone=GMT%2B8&autoReconnect=true&rewriteBatchedStatements=true&allowPublicKeyRetrieval=true&nullCatalogMeansCurrent=true
username: root
password: hn02le.34lkdLKD
# # 从库数据源
# slave:
# lazy: true
# type: ${spring.datasource.type}
# driverClassName: com.mysql.cj.jdbc.Driver
# url: jdbc:mysql://localhost:3306/ry-vue?useUnicode=true&characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=true&serverTimezone=GMT%2B8&autoReconnect=true&rewriteBatchedStatements=true&allowPublicKeyRetrieval=true&nullCatalogMeansCurrent=true
# username:
# password:
# oracle:
# type: ${spring.datasource.type}
# driverClassName: oracle.jdbc.OracleDriver
# url: jdbc:oracle:thin:@//localhost:1521/XE
# username: ROOT
# password: root
# postgres:
# type: ${spring.datasource.type}
# driverClassName: org.postgresql.Driver
# url: jdbc:postgresql://localhost:5432/postgres?useUnicode=true&characterEncoding=utf8&useSSL=true&autoReconnect=true&reWriteBatchedInserts=true
# username: root
# password: root
# sqlserver:
# type: ${spring.datasource.type}
# driverClassName: com.microsoft.sqlserver.jdbc.SQLServerDriver
# url: jdbc:sqlserver://localhost:1433;DatabaseName=tempdb;SelectMethod=cursor;encrypt=false;rewriteBatchedStatements=true
# username: SA
# password: root
hikari:
# 最大连接池数量
maxPoolSize: 30
# 最小空闲线程数量
minIdle: 10
# 配置获取连接等待超时的时间
connectionTimeout: 30000
# 校验超时时间
validationTimeout: 5000
# 空闲连接存活最大时间,默认10分钟
idleTimeout: 600000
# 此属性控制池中连接的最长生命周期,值0表示无限生命周期,默认30分钟
maxLifetime: 1800000
# 多久检查一次连接的活性
keepaliveTime: 30000
--- # redis 单机配置(单机与集群只能开启一个另一个需要注释掉)
spring.data:
redis:
# 地址
host: 47.99.164.66
# 端口,默认为6379
port: 16379
# 数据库索引
database: 3
# redis 密码必须配置
password: Hzdyjs@123456
# 连接超时时间
timeout: 10s
# 是否开启ssl
ssl.enabled: false
# redisson 配置
redisson:
# redis key前缀
keyPrefix:
# 线程池数量
threads: 16
# Netty线程池数量
nettyThreads: 32
# 单节点配置
singleServerConfig:
# 客户端名称 不能用中文
clientName: RuoYi-Vue-Plus
# 最小空闲连接数
connectionMinimumIdleSize: 32
# 连接池大小
connectionPoolSize: 64
# 连接空闲超时,单位:毫秒
idleConnectionTimeout: 10000
# 命令等待超时,单位:毫秒
timeout: 3000
# 发布和订阅连接池大小
subscriptionConnectionPoolSize: 50
--- # mail 邮件发送
mail:
enabled: false
host: smtp.163.com
port: 465
# 是否需要用户名密码验证
auth: true
# 发送方,遵循RFC-822标准
from: xxx@163.com
# 用户名(注意:如果使用foxmail邮箱,此处user为qq号)
user: xxx@163.com
# 密码(注意,某些邮箱需要为SMTP服务单独设置密码,详情查看相关帮助)
pass: xxxxxxxxxx
# 使用 STARTTLS安全连接,STARTTLS是对纯文本通信协议的扩展。
starttlsEnable: true
# 使用SSL安全连接
sslEnable: true
# SMTP超时时长,单位毫秒,缺省值不超时
timeout: 0
# Socket连接超时值,单位毫秒,缺省值不超时
connectionTimeout: 0
--- # sms 短信 支持 阿里云 腾讯云 云片 等等各式各样的短信服务商
# https://sms4j.com/doc3/ 差异配置文档地址 支持单厂商多配置,可以配置多个同时使用
sms:
# 配置源类型用于标定配置来源(interface,yaml)
config-type: yaml
# 用于标定yml中的配置是否开启短信拦截,接口配置不受此限制
restricted: true
# 短信拦截限制单手机号每分钟最大发送,只对开启了拦截的配置有效
minute-max: 1
# 短信拦截限制单手机号每日最大发送量,只对开启了拦截的配置有效
account-max: 30
# 以下配置来自于 org.dromara.sms4j.provider.config.BaseConfig类中
blends:
# 唯一ID 用于发送短信寻找具体配置 随便定义别用中文即可
# 可以同时存在两个相同厂商 例如: ali1 ali2 两个不同的阿里短信账号 也可用于区分租户
config1:
# 框架定义的厂商名称标识,标定此配置是哪个厂商,详细请看厂商标识介绍部分
supplier: alibaba
# 有些称为accessKey有些称之为apiKey,也有称为sdkKey或者appId。
access-key-id: 您的accessKey
# 称为accessSecret有些称之为apiSecret
access-key-secret: 您的accessKeySecret
signature: 您的短信签名
sdk-app-id: 您的sdkAppId
config2:
# 厂商标识,标定此配置是哪个厂商,详细请看厂商标识介绍部分
supplier: tencent
access-key-id: 您的accessKey
access-key-secret: 您的accessKeySecret
signature: 您的短信签名
sdk-app-id: 您的sdkAppId
--- # 三方授权
justauth:
# 前端外网访问地址
address: http://localhost:80
type:
maxkey:
# maxkey 服务器地址
# 注意 如下均配置均不需要修改 maxkey 已经内置好了数据
server-url: http://sso.maxkey.top
client-id: 876892492581044224
client-secret: x1Y5MTMwNzIwMjMxNTM4NDc3Mzche8
redirect-uri: ${justauth.address}/social-callback?source=maxkey
topiam:
# topiam 服务器地址
server-url: http://127.0.0.1:1989/api/v1/authorize/y0q************spq***********8ol
client-id: 449c4*********937************759
client-secret: ac7***********1e0************28d
redirect-uri: ${justauth.address}/social-callback?source=topiam
scopes: [ openid, email, phone, profile ]
qq:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=qq
union-id: false
weibo:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=weibo
gitee:
client-id: 91436b7940090d09c72c7daf85b959cfd5f215d67eea73acbf61b6b590751a98
client-secret: 02c6fcfd70342980cd8dd2f2c06c1a350645d76c754d7a264c4e125f9ba915ac
redirect-uri: ${justauth.address}/social-callback?source=gitee
dingtalk:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=dingtalk
baidu:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=baidu
csdn:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=csdn
coding:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=coding
coding-group-name: xx
oschina:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=oschina
alipay_wallet:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=alipay_wallet
alipay-public-key: MIIB**************DAQAB
wechat_open:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=wechat_open
wechat_mp:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=wechat_mp
wechat_enterprise:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=wechat_enterprise
agent-id: 1000002
gitlab:
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=gitlab
gitea:
# 前端改动 https://gitee.com/JavaLionLi/plus-ui/pulls/204
# gitea 服务器地址
server-url: https://demo.gitea.com
client-id: 10**********6
client-secret: 1f7d08**********5b7**********29e
redirect-uri: ${justauth.address}/social-callback?source=gitea
system:
file-cache:
# 是否启用文件缓存
enabled: true
# 最大缓存文件数量
max-cache-size: 100
# 最大缓存大小(字节)- 500MB
max-cache-size-bytes: 524288000
# 缓存过期时间(毫秒)- 30分钟
expire-time-ms: 1800000
# 最大缓存单个文件大小(字节)- 50MB
max-single-file-size: 52428800
# LRU淘汰比例(百分比)
evict-ratio: 20
# 是否启用缓存统计
enable-stats: true
# 缓存清理间隔(毫秒)- 5分钟
cleanup-interval-ms: 300000
# 配置说明:
# 1. enabled: 控制是否启用文件缓存功能
# 2. max-cache-size: 缓存的最大文件数量
# 3. max-cache-size-bytes: 缓存的最大总大小(字节)
# 4. expire-time-ms: 缓存条目的过期时间(毫秒)
# 5. max-single-file-size: 单个文件的最大缓存大小(字节)
# 6. evict-ratio: LRU淘汰时移除的条目比例(百分比)
# 7. enable-stats: 是否启用缓存统计功能
# 8. cleanup-interval-ms: 自动清理过期条目的间隔(毫秒)
# 性能优化建议:
# - 对于内存充足的环境,可以适当增加 max-cache-size-bytes
# - 对于频繁访问的小文件,可以增加 max-cache-size
# - 对于大文件较多的环境,可以调整 max-single-file-size
# - 对于访问频率较低的环境,可以适当减少 expire-time-ms
--- # SQL日志配置(用于调试班级名额查询问题)
logging:
level:
# 开启 MyBatis SQL 日志
org.dromara.jcbt.mapper: DEBUG
org.dromara.jcbt.service: DEBUG
# 开启 MyBatis-Plus SQL 日志
com.baomidou.mybatisplus.extension.plugins.MybatisPlusInterceptor: DEBUG
# 开启 HikariCP 连接池日志(可选)
com.zaxxer.hikari: DEBUG
# 开启 P6Spy SQL 日志(更详细的SQL执行日志)
p6spy: INFO
pattern:
# 控制台日志格式,增加线程名和方法名
console: "%clr(%d{yyyy-MM-dd HH:mm:ss.SSS}){faint} %clr(%5p) %clr(${PID:- }){magenta} %clr(---){faint} %clr([%15.15t]){faint} %clr(%-40.40logger{39}){cyan} %clr(:){faint} %m%n%wEx"
# MyBatis 配置
mybatis-plus:
configuration:
# 开启 SQL 日志打印
log-impl: org.apache.ibatis.logging.stdout.StdOutImpl
# 开启驼峰命名转换
map-underscore-to-camel-case: true
# 开启缓存
cache-enabled: true
# 开启懒加载
lazy-loading-enabled: true
# 开启积极懒加载
aggressive-lazy-loading: false
--- # 高德地图配置
amap:
# 高德地图API密钥(请替换为您的实际密钥)
key: 920750e421cd6357ddf60ee1e929e0dd
# 高德地图API基础URL
url: https://restapi.amap.com
@@ -5,7 +5,7 @@ spring.servlet.multipart.location: /ruoyi/server/temp
spring.boot.admin.client:
# 增加客户端开关
enabled: true
url: http://47.121.125.41:9090/admin
url: http://172.16.0.67:9090/admin
instance:
service-host-type: IP
metadata:
@@ -22,7 +22,7 @@ snail-job:
# SnailJob 接入验证令牌 详见 script/sql/ry_job.sql `sj_group_config`表
token: "SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT"
server:
host: 47.121.125.41
host: 172.16.0.67
port: 17888
# 命名空间UUID 详见 script/sql/ry_job.sql `sj_namespace`表`unique_id`字段
namespace: ${spring.profiles.active}
@@ -52,9 +52,9 @@ spring:
driverClassName: com.mysql.cj.jdbc.Driver
# jdbc 所有参数配置参考 https://lionli.blog.csdn.net/article/details/122018562
# rewriteBatchedStatements=true 批处理优化 大幅提升批量插入更新删除性能(对数据库有性能损耗 使用批量操作应考虑性能问题)
url: jdbc:mysql://47.99.164.66:3306/hzdyjs_xg?useUnicode=true&characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=true&serverTimezone=GMT%2B8&autoReconnect=true&rewriteBatchedStatements=true&allowPublicKeyRetrieval=true&nullCatalogMeansCurrent=true
url: jdbc:mysql://172.16.0.67:3306/hzdyjs_xg?useUnicode=true&characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=true&serverTimezone=GMT%2B8&autoReconnect=true&rewriteBatchedStatements=true&allowPublicKeyRetrieval=true&nullCatalogMeansCurrent=true
username: root
password: hn02le.34lkdLKD
password: Hzdyjs@Root@123456
# # 从库数据源
# slave:
# lazy: true
@@ -83,7 +83,7 @@ spring:
# password: root
hikari:
# 最大连接池数量
maxPoolSize: 30
maxPoolSize: 60
# 最小空闲线程数量
minIdle: 10
# 配置获取连接等待超时的时间
@@ -94,20 +94,20 @@ spring:
idleTimeout: 600000
# 此属性控制池中连接的最长生命周期,值0表示无限生命周期,默认30分钟
maxLifetime: 1800000
# 多久检查一次连接的活性
keepaliveTime: 30000
# 多久检查一次连接的活性(增加间隔,减少检测频率,减少日志输出)
keepaliveTime: 300000
--- # redis 单机配置(单机与集群只能开启一个另一个需要注释掉)
spring.data:
redis:
# 地址
host: 47.99.164.66
host: 172.16.0.67
# 端口,默认为6379
port: 16379
# 数据库索引
database: 3
# redis 密码必须配置
password: Hzdyjs@123456
password: Hzdyjs@redis@123456
# 连接超时时间
timeout: 10s
# 是否开启ssl
@@ -318,15 +318,16 @@ system:
--- # SQL日志配置(用于调试班级名额查询问题)
logging:
level:
# 开启 MyBatis SQL 日志
org.dromara.jcbt.mapper: DEBUG
org.dromara.jcbt.service: DEBUG
# 开启 MyBatis-Plus SQL 日志
com.baomidou.mybatisplus.extension.plugins.MybatisPlusInterceptor: DEBUG
# 开启 HikariCP 连接池日志(可选)
com.zaxxer.hikari: DEBUG
# 开启 P6Spy SQL 日志(更详细的SQL执行日志)
p6spy: INFO
# 生产环境关闭 MyBatis SQL 日志,避免性能影响和敏感信息泄露
# org.dromara.jcbt.mapper: DEBUG
# org.dromara.jcbt.service: DEBUG
# 生产环境关闭 MyBatis-Plus SQL 日志
# com.baomidou.mybatisplus.extension.plugins.MybatisPlusInterceptor: DEBUG
# 关闭 HikariCP 连接池DEBUG日志,减少keepalive检测日志
com.zaxxer.hikari: WARN
com.zaxxer.hikari.pool.HikariPool: WARN
# 生产环境关闭 P6Spy SQL 日志,避免性能影响
# p6spy: INFO
pattern:
# 控制台日志格式,增加线程名和方法名
console: "%clr(%d{yyyy-MM-dd HH:mm:ss.SSS}){faint} %clr(%5p) %clr(${PID:- }){magenta} %clr(---){faint} %clr([%15.15t]){faint} %clr(%-40.40logger{39}){cyan} %clr(:){faint} %m%n%wEx"
@@ -334,7 +335,7 @@ logging:
# MyBatis 配置
mybatis-plus:
configuration:
# 开启 SQL 日志打印
# 生产环境关闭 SQL 日志打印,避免性能影响和敏感信息泄露
log-impl: org.apache.ibatis.logging.stdout.StdOutImpl
# 开启驼峰命名转换
map-underscore-to-camel-case: true
@@ -345,7 +346,7 @@ mybatis-plus:
# 开启积极懒加载
aggressive-lazy-loading: false
--- # 高德地图配置
--- # 高德地图配置 生产环境需要修改成学校的配置
amap:
# 高德地图API密钥(请替换为您的实际密钥)
key: 920750e421cd6357ddf60ee1e929e0dd
@@ -41,6 +41,9 @@ logging:
org.springframework: warn
org.mybatis.spring.mapper: error
org.apache.fury: warn
# 关闭HikariCP连接池DEBUG日志,减少keepalive检测日志
com.zaxxer.hikari: warn
com.zaxxer.hikari.pool.HikariPool: warn
# 关闭SQL日志调试,减少SELECT 1心跳检测日志
# org.dromara.cw.mapper.CwApprovalRecordMapper: info
# org.dromara.cw.service.impl.CwApprovalRecordServiceImpl: info
@@ -165,4 +165,9 @@
<appender-ref ref="file_sql" />
</logger>
<!-- 关闭HikariCP连接池的DEBUG日志,减少keepalive检测日志 -->
<logger name="com.zaxxer.hikari" level="WARN"/>
<logger name="com.zaxxer.hikari.pool.HikariPool" level="WARN"/>
<logger name="com.zaxxer.hikari.HikariConfig" level="WARN"/>
</configuration>
@@ -38,7 +38,7 @@ spring.boot.admin.client:
# 增加客户端开关
enabled: true
# 设置 Spring Boot Admin Server 地址
url: http://127.0.0.1:9090/admin
url: http://172.16.0.67:9090/admin
instance:
service-host-type: IP
metadata:
@@ -0,0 +1,40 @@
spring:
datasource:
type: com.zaxxer.hikari.HikariDataSource
driver-class-name: com.mysql.cj.jdbc.Driver
url: jdbc:mysql://47.99.164.66:3306/hzdyjs_xg?useUnicode=true&characterEncoding=UTF-8&serverTimezone=Asia/Shanghai&allowPublicKeyRetrieval=true
username: root
password: hn02le.34lkdLKD
hikari:
connection-timeout: 30000
validation-timeout: 5000
minimum-idle: 10
maximum-pool-size: 20
idle-timeout: 600000
max-lifetime: 900000
keepaliveTime: 30000
--- # snail-job 服务端配置
snail-job:
# 拉取重试数据的每批次的大小
retry-pull-page-size: 1000
# 拉取重试数据的每批次的大小
job-pull-page-size: 1000
# 服务器端口
server-port: 17888
# 日志保存时间(单位: day)
log-storage: 7
rpc-type: grpc
--- # 监控中心配置
spring.boot.admin.client:
# 增加客户端开关
enabled: true
url: http://47.121.125.41:9090/admin
instance:
service-host-type: IP
metadata:
username: ${spring.boot.admin.client.username}
userpassword: ${spring.boot.admin.client.password}
username: @monitor.username@
password: @monitor.password@
@@ -2,9 +2,9 @@ spring:
datasource:
type: com.zaxxer.hikari.HikariDataSource
driver-class-name: com.mysql.cj.jdbc.Driver
url: jdbc:mysql://47.99.164.66:3306/hzdyjs_xg?useUnicode=true&characterEncoding=UTF-8&serverTimezone=Asia/Shanghai&allowPublicKeyRetrieval=true
url: jdbc:mysql://172.16.0.67:3306/hzdyjs_xg?useUnicode=true&characterEncoding=UTF-8&serverTimezone=Asia/Shanghai&allowPublicKeyRetrieval=true
username: root
password: hn02le.34lkdLKD
password: Hzdyjs@Root@123456
hikari:
connection-timeout: 30000
validation-timeout: 5000
@@ -30,7 +30,7 @@ snail-job:
spring.boot.admin.client:
# 增加客户端开关
enabled: true
url: http://47.121.125.41:9090/admin
url: http://172.16.0.67:9090/admin
instance:
service-host-type: IP
metadata:
@@ -0,0 +1,252 @@
# SSE关闭接口406错误解决方案
## 🚨 问题描述
用户在前台退出时,SSE关闭接口返回406 Not Acceptable错误:
```
Request URL: http://hzdyjsweb.camasort.com/prod-api/resource/sse/close
Request Method: GET
Status Code: 406 Not Acceptable
```
## 🔍 问题分析
### 1. 错误原因分析
**406 Not Acceptable错误**通常表示:
- 服务器无法根据请求的内容特性完成请求
- 请求的资源的内容特性无法满足请求头中的条件
- 用户认证信息缺失或无效
### 2. 根本原因
1. **认证时机问题**:
- 前端在logout时先调用SSE关闭接口,再调用logout接口
- 此时token可能已经失效或即将失效
- 后端`LoginHelper.getUserId()`可能返回null
2. **后端接口设计问题**:
- SSE关闭接口使用了`@SaIgnore`注解,但仍然需要获取用户信息
- 缺少对空值和异常的处理
- 没有优雅降级机制
3. **nginx代理配置问题**:
- SSE关闭接口可能没有正确的代理配置
- 缺少专门的错误处理
## 🛠️ 解决方案
### 方案一:修复后端SSE关闭接口(已实施)
#### 1. 优化错误处理
```java
/**
* 关闭 SSE 连接
*/
@SaIgnore
@GetMapping(value = "${sse.path}/close")
public R<Void> close() {
try {
// 尝试获取token和用户ID,如果获取失败则直接返回成功
String tokenValue = null;
Long userId = null;
try {
tokenValue = StpUtil.getTokenValue();
userId = LoginHelper.getUserId();
} catch (Exception e) {
// 如果获取用户信息失败,记录日志但不抛出异常
log.warn("Failed to get user info for SSE close: {}", e.getMessage());
return R.ok();
}
// 如果token或userId为空,直接返回成功
if (tokenValue == null || userId == null) {
log.info("Token or userId is null, skipping SSE disconnect");
return R.ok();
}
sseEmitterManager.disconnect(userId, tokenValue);
return R.ok();
} catch (Exception e) {
log.error("Error during SSE close: {}", e.getMessage(), e);
// 即使出错也返回成功,避免影响用户退出流程
return R.ok();
}
}
```
#### 2. 关键改进点
- **添加异常捕获**:捕获所有可能的异常
- **空值检查**:检查token和userId是否为空
- **优雅降级**:即使出错也返回成功状态
- **日志记录**:记录详细的错误信息用于调试
### 方案二:优化前端logout逻辑(已实施)
#### 1. 添加错误处理
```typescript
/**
* 注销
*/
export function logout() {
// 先尝试关闭SSE连接,但不阻塞logout流程
if (import.meta.env.VITE_APP_SSE === 'true') {
request({
url: '/resource/sse/close',
method: 'get'
}).catch((error) => {
// SSE关闭失败不影响用户退出,只记录日志
console.warn('SSE close failed during logout:', error);
});
}
// 执行正常的logout流程
return request({
url: '/auth/logout',
method: 'post'
});
}
```
#### 2. 关键改进点
- **非阻塞调用**:SSE关闭失败不影响logout流程
- **错误捕获**:捕获SSE关闭接口的错误
- **日志记录**:记录错误信息但不影响用户体验
### 方案三:更新nginx配置(已实施)
#### 1. 添加SSE关闭接口专用配置
```nginx
# SSE关闭接口专用配置
location ^~ /prod-api/resource/sse/close {
proxy_pass http://47.121.125.41:8080/resource/sse/close;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
# 普通API配置(不需要SSE特殊配置)
proxy_connect_timeout 30s;
proxy_send_timeout 30s;
proxy_read_timeout 30s;
# CORS处理
add_header Access-Control-Allow-Origin *;
add_header Access-Control-Allow-Methods 'GET, POST, PUT, DELETE, OPTIONS';
add_header Access-Control-Allow-Headers 'DNT,User-Agent,X-Requested-With,If-Modified-Since,Cache-Control,Content-Type,Range,Authorization,clientid';
add_header Access-Control-Allow-Credentials true;
if ($request_method = 'OPTIONS') {
return 204;
}
}
```
#### 2. 配置说明
- **独立配置**:为SSE关闭接口提供独立的代理配置
- **合理超时**:设置30秒超时,适合普通API调用
- **CORS支持**:确保跨域请求正常工作
## 📋 部署步骤
### 1. 更新后端代码
```bash
# 重新编译后端项目
mvn clean compile
# 重启后端服务
systemctl restart your-backend-service
```
### 2. 更新前端代码
```bash
# 重新构建前端项目
npm run build
# 部署到nginx目录
cp -r dist/* /root/hzdyjsweb9091/dist/
```
### 3. 更新nginx配置
```bash
# 测试nginx配置
nginx -t
# 重新加载配置
nginx -s reload
```
## 🔧 验证方法
### 1. 功能验证
1. **正常登录**:确保用户能够正常登录
2. **SSE连接**:检查SSE连接是否正常建立
3. **正常退出**:测试用户退出流程
4. **错误处理**:检查控制台是否有错误信息
### 2. 错误场景测试
1. **Token失效退出**:在token即将失效时退出
2. **网络异常退出**:在网络不稳定的情况下退出
3. **并发退出**:多个用户同时退出
### 3. 日志检查
```bash
# 检查后端日志
tail -f /path/to/backend/logs/application.log | grep -i "sse"
# 检查nginx错误日志
tail -f /var/log/nginx/hzdyjsweb.camasort.com.error.log
```
## 🎯 预期效果
### 1. 问题解决
- ✅ 消除406错误
- ✅ 用户退出流程顺畅
- ✅ 不影响用户体验
### 2. 系统稳定性
- ✅ SSE连接正确关闭
- ✅ 资源正确释放
- ✅ 错误处理完善
### 3. 监控指标
- SSE关闭成功率:100%
- 用户退出成功率:100%
- 错误日志减少:显著减少
## 📚 相关文档
- [SSE生产环境配置问题解决方案](../SSE生产环境配置问题解决方案.md)
- [nginx配置文档](../../部署文档/nginx配置/)
- [前端环境变量配置建议](../../部署文档/nginx配置/前端环境变量配置建议.md)
## 🔄 后续优化
1. **监控告警**:添加SSE关闭失败监控
2. **性能优化**:优化SSE连接管理
3. **用户体验**:添加退出进度提示
4. **错误恢复**:实现自动重试机制
---
**更新时间**:2025-01-03
**维护人员**:系统管理员
**文档版本**:v1.0
@@ -0,0 +1,256 @@
# SSE生产环境配置问题解决方案
## 🚨 问题描述
在生产环境中,通过nginx配置后,前端无法正常接收SSE(Server-Sent Events)消息推送。
## 🔍 问题分析
### 1. 技术架构分析
**后端SSE实现**:
- 使用Spring Boot的`SseEmitter`实现
- SSE端点:`/resource/sse`
- 支持用户认证和Token验证
- 使用Redis进行消息广播
**前端SSE实现**:
- 使用Vue 3的`useEventSource`组合式API
- 自动重连机制(10次重试,3秒间隔)
- 消息处理:显示通知和存储到状态管理
**nginx代理配置**:
- 当前配置缺少专门的SSE代理处理
- 超时设置可能不够长
- 缺少SSE特定的响应头配置
### 2. 根本原因
1. **nginx配置不完整**:缺少专门的SSE代理配置
2. **超时设置不足**:SSE是长连接,需要更长的超时时间
3. **响应头配置缺失**:SSE需要特定的Content-Type和缓存控制
4. **代理优先级问题**:SSE路由可能被其他配置覆盖
## 🛠️ 解决方案
### 方案一:优化nginx配置(推荐)
#### 1. 添加专门的SSE代理配置
```nginx
# SSE专用代理配置(重要:必须放在其他API配置之前)
location ^~ /prod-api/resource/sse {
proxy_pass http://47.121.125.41:8080/resource/sse;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
# SSE特殊配置
proxy_http_version 1.1;
proxy_set_header Connection "";
proxy_buffering off;
proxy_cache off;
# SSE需要更长的超时时间
proxy_connect_timeout 60s;
proxy_send_timeout 86400s; # 24小时
proxy_read_timeout 86400s; # 24小时
# SSE响应头设置
proxy_set_header Accept "text/event-stream";
proxy_set_header Cache-Control "no-cache";
proxy_set_header Connection "keep-alive";
# CORS处理
add_header Access-Control-Allow-Origin *;
add_header Access-Control-Allow-Methods 'GET, POST, PUT, DELETE, OPTIONS';
add_header Access-Control-Allow-Headers 'DNT,User-Agent,X-Requested-With,If-Modified-Since,Cache-Control,Content-Type,Range,Authorization,clientid';
add_header Access-Control-Allow-Credentials true;
if ($request_method = 'OPTIONS') {
return 204;
}
}
```
#### 2. 关键配置说明
- **`proxy_buffering off`**:禁用nginx缓冲,确保SSE消息实时传输
- **`proxy_cache off`**:禁用缓存,避免消息被缓存
- **`proxy_read_timeout 86400s`**:设置24小时超时,支持长连接
- **`proxy_set_header Connection ""`**:清除Connection头,避免代理问题
- **`add_header Access-Control-Allow-Credentials true`**:支持跨域认证
### 方案二:修改前端配置
#### 1. 环境变量配置
```bash
# 推荐配置:通过nginx代理
VITE_APP_BASE_API = '/prod-api'
VITE_APP_SSE = true
VITE_APP_WEBSOCKET = false
```
#### 2. 备选配置:直接访问后端
```bash
# 备选配置:直接访问后端
VITE_APP_BASE_API = 'http://47.121.125.41:8080'
VITE_APP_SSE = true
VITE_APP_WEBSOCKET = false
```
## 📋 部署步骤
### 1. 更新nginx配置
```bash
# 备份原配置
cp /etc/nginx/conf.d/hzdyjsweb.camasort.com.conf /etc/nginx/conf.d/hzdyjsweb.camasort.com.conf.backup
# 更新配置文件
vim /etc/nginx/conf.d/hzdyjsweb.camasort.com.conf
# 测试配置
nginx -t
# 重新加载配置
nginx -s reload
```
### 2. 更新前端配置
```bash
# 进入前端项目目录
cd /path/to/frontend
# 更新环境变量
vim .env.production
# 重新构建
npm run build
# 部署到nginx目录
cp -r dist/* /root/hzdyjsweb9091/dist/
```
### 3. 重启服务
```bash
# 重启后端服务(如果需要)
systemctl restart your-backend-service
# 检查nginx状态
systemctl status nginx
```
## 🔧 故障排查
### 1. 检查nginx配置
```bash
# 测试nginx配置
nginx -t
# 查看nginx错误日志
tail -f /var/log/nginx/error.log
tail -f /var/log/nginx/hzdyjsweb.camasort.com.error.log
# 查看nginx访问日志
tail -f /var/log/nginx/hzdyjsweb.camasort.com.access.log
```
### 2. 检查SSE连接
#### 浏览器开发者工具
1. 打开F12开发者工具
2. 切换到Network面板
3. 筛选XHR/Fetch请求
4. 查找SSE连接(通常显示为EventStream)
5. 检查连接状态和响应头
#### 控制台错误信息
```javascript
// 检查SSE连接状态
console.log('SSE连接状态:', EventSource.readyState);
// 0: CONNECTING, 1: OPEN, 2: CLOSED
```
### 3. 测试SSE端点
#### 直接测试后端
```bash
curl -H "Authorization: Bearer YOUR_TOKEN" \
-H "clientid: e5cd7e4891bf95d1d19206ce24a7b32e" \
http://47.121.125.41:8080/resource/sse
```
#### 通过nginx测试
```bash
curl -H "Authorization: Bearer YOUR_TOKEN" \
-H "clientid: e5cd7e4891bf95d1d19206ce24a7b32e" \
http://hzdyjsweb.camasort.com/prod-api/resource/sse
```
### 4. 检查后端SSE服务
#### 查看后端日志
```bash
# 查看应用日志
tail -f /path/to/backend/logs/application.log
# 查看SSE相关日志
grep -i "sse" /path/to/backend/logs/application.log
```
#### 检查Redis连接
```bash
# 连接Redis
redis-cli
# 检查SSE主题订阅
SUBSCRIBE global:sse
```
## 🎯 验证方法
### 1. 功能验证
1. **登录系统**:确保用户已登录
2. **触发SSE消息**:通过工作流或其他功能触发消息推送
3. **检查通知**:查看是否收到实时通知
4. **检查控制台**:查看SSE连接状态和消息
### 2. 性能验证
1. **连接稳定性**:长时间保持连接,检查是否断开
2. **消息延迟**:测试消息推送的实时性
3. **并发连接**:测试多用户同时连接的情况
### 3. 监控指标
- SSE连接数
- 消息推送成功率
- 连接断开频率
- 响应时间
## 📚 相关文档
- [nginx SSE代理配置官方文档](http://nginx.org/en/docs/http/ngx_http_proxy_module.html)
- [Spring Boot SSE官方文档](https://docs.spring.io/spring-framework/reference/web/webmvc/controller/ann-methods/sse.html)
- [Vue 3 EventSource使用指南](https://vueuse.org/core/useEventSource/)
## 🔄 后续优化
1. **添加SSE连接监控**:监控连接状态和性能
2. **实现连接池管理**:优化连接资源使用
3. **添加消息重试机制**:提高消息推送可靠性
4. **优化nginx配置**:根据实际使用情况调整超时和缓冲设置
---
**更新时间**:2025-01-03
**维护人员**:系统管理员
**文档版本**:v1.0
@@ -41,36 +41,95 @@ server {
add_header Pragma "no-cache";
}
# Warm-Flow工作流UI代理配置(重要:必须放在Vue路由配置之前,使用^~确保优先级)
location ^~ /prod-api/warm-flow-ui/ {
proxy_pass http://47.121.125.41:8080/warm-flow-ui/;
# SSE专用代理配置(重要:必须放在其他API配置之前)
location ^~ /prod-api/resource/sse {
proxy_pass http://47.121.125.41:8080/resource/sse;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
# 支持WebSocket和SSE
# SSE特殊配置
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
proxy_set_header Connection "";
proxy_buffering off;
proxy_cache off;
# 超时设置
# SSE需要更长的超时时间
proxy_connect_timeout 60s;
proxy_send_timeout 60s;
proxy_read_timeout 86400s;
proxy_send_timeout 86400s; # 24小时
proxy_read_timeout 86400s; # 24小时
# SSE响应头设置
proxy_set_header Accept "text/event-stream";
proxy_set_header Cache-Control "no-cache";
proxy_set_header Connection "keep-alive";
# CORS处理
add_header Access-Control-Allow-Origin *;
add_header Access-Control-Allow-Methods 'GET, POST, PUT, DELETE, OPTIONS';
add_header Access-Control-Allow-Headers 'DNT,User-Agent,X-Requested-With,If-Modified-Since,Cache-Control,Content-Type,Range,Authorization,clientid';
add_header Access-Control-Allow-Credentials true;
if ($request_method = 'OPTIONS') {
return 204;
}
}
# SSE关闭接口专用配置
location ^~ /prod-api/resource/sse/close {
proxy_pass http://47.121.125.41:8080/resource/sse/close;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
# 普通API配置(不需要SSE特殊配置)
proxy_connect_timeout 30s;
proxy_send_timeout 30s;
proxy_read_timeout 30s;
# CORS处理
add_header Access-Control-Allow-Origin *;
add_header Access-Control-Allow-Methods 'GET, POST, PUT, DELETE, OPTIONS';
add_header Access-Control-Allow-Headers 'DNT,User-Agent,X-Requested-With,If-Modified-Since,Cache-Control,Content-Type,Range,Authorization,clientid';
add_header Access-Control-Allow-Credentials true;
if ($request_method = 'OPTIONS') {
return 204;
}
}
# Warm-Flow工作流UI代理配置(重要:必须放在Vue路由配置之前)
location ^~ /prod-api/warm-flow-ui/ {
proxy_pass http://47.121.125.41:8080/warm-flow-ui/;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
# 支持WebSocket和SSE
proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
proxy_buffering off;
proxy_cache off;
# 超时设置
proxy_connect_timeout 60s;
proxy_send_timeout 60s;
proxy_read_timeout 86400s;
# CORS处理
add_header Access-Control-Allow-Origin *;
add_header Access-Control-Allow-Methods 'GET, POST, PUT, DELETE, OPTIONS';
add_header Access-Control-Allow-Headers 'DNT,User-Agent,X-Requested-With,If-Modified-Since,Cache-Control,Content-Type,Range,Authorization,clientid';
if ($request_method = 'OPTIONS') {
return 204;
}
}
# API代理配置(使用^~确保优先级,避免被Vue路由配置干扰)
location ^~ /prod-api/ {
proxy_pass http://47.121.125.41:8080/;
@@ -0,0 +1,135 @@
version: '3.8'
services:
# 9900端口实例
ruoyi-admin-9900:
image: openjdk:17-jre-slim
container_name: ruoyi-admin-9900
ports:
- "9900:9900"
volumes:
- ./ruoyi-admin.jar:/app/ruoyi-admin.jar
- ./logs:/app/logs
- ./temp:/app/temp
working_dir: /app
command: >
java -Xms512m -Xmx1024m
-XX:MetaspaceSize=128m -XX:MaxMetaspaceSize=256m
-Dfile.encoding=UTF-8 -Duser.timezone=Asia/Shanghai
-XX:+UseG1GC -XX:MaxGCPauseMillis=200
-XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=logs/
-Dspring.profiles.active=prod-9900
-Dserver.port=9900
-jar ruoyi-admin.jar
environment:
- SPRING_PROFILES_ACTIVE=prod-9900
- SERVER_PORT=9900
restart: unless-stopped
networks:
- ruoyi-network
# 9901端口实例
ruoyi-admin-9901:
image: openjdk:17-jre-slim
container_name: ruoyi-admin-9901
ports:
- "9901:9901"
volumes:
- ./ruoyi-admin.jar:/app/ruoyi-admin.jar
- ./logs:/app/logs
- ./temp:/app/temp
working_dir: /app
command: >
java -Xms512m -Xmx1024m
-XX:MetaspaceSize=128m -XX:MaxMetaspaceSize=256m
-Dfile.encoding=UTF-8 -Duser.timezone=Asia/Shanghai
-XX:+UseG1GC -XX:MaxGCPauseMillis=200
-XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=logs/
-Dspring.profiles.active=prod-9901
-Dserver.port=9901
-jar ruoyi-admin.jar
environment:
- SPRING_PROFILES_ACTIVE=prod-9901
- SERVER_PORT=9901
restart: unless-stopped
networks:
- ruoyi-network
# 9902端口实例
ruoyi-admin-9902:
image: openjdk:17-jre-slim
container_name: ruoyi-admin-9902
ports:
- "9902:9902"
volumes:
- ./ruoyi-admin.jar:/app/ruoyi-admin.jar
- ./logs:/app/logs
- ./temp:/app/temp
working_dir: /app
command: >
java -Xms512m -Xmx1024m
-XX:MetaspaceSize=128m -XX:MaxMetaspaceSize=256m
-Dfile.encoding=UTF-8 -Duser.timezone=Asia/Shanghai
-XX:+UseG1GC -XX:MaxGCPauseMillis=200
-XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=logs/
-Dspring.profiles.active=prod-9902
-Dserver.port=9902
-jar ruoyi-admin.jar
environment:
- SPRING_PROFILES_ACTIVE=prod-9902
- SERVER_PORT=9902
restart: unless-stopped
networks:
- ruoyi-network
# 9903端口实例
ruoyi-admin-9903:
image: openjdk:17-jre-slim
container_name: ruoyi-admin-9903
ports:
- "9903:9903"
volumes:
- ./ruoyi-admin.jar:/app/ruoyi-admin.jar
- ./logs:/app/logs
- ./temp:/app/temp
working_dir: /app
command: >
java -Xms512m -Xmx1024m
-XX:MetaspaceSize=128m -XX:MaxMetaspaceSize=256m
-Dfile.encoding=UTF-8 -Duser.timezone=Asia/Shanghai
-XX:+UseG1GC -XX:MaxGCPauseMillis=200
-XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=logs/
-Dspring.profiles.active=prod-9903
-Dserver.port=9903
-jar ruoyi-admin.jar
environment:
- SPRING_PROFILES_ACTIVE=prod-9903
- SERVER_PORT=9903
restart: unless-stopped
networks:
- ruoyi-network
# Nginx负载均衡器
nginx:
image: nginx:alpine
container_name: ruoyi-nginx
ports:
- "80:80"
volumes:
- ./nginx.conf:/etc/nginx/nginx.conf:ro
depends_on:
- ruoyi-admin-9900
- ruoyi-admin-9901
- ruoyi-admin-9902
- ruoyi-admin-9903
restart: unless-stopped
networks:
- ruoyi-network
networks:
ruoyi-network:
driver: bridge
volumes:
logs:
temp:
@@ -0,0 +1,116 @@
events {
worker_connections 1024;
}
http {
include /etc/nginx/mime.types;
default_type application/octet-stream;
# 日志格式
log_format main '$remote_addr - $remote_user [$time_local] "$request" '
'$status $body_bytes_sent "$http_referer" '
'"$http_user_agent" "$http_x_forwarded_for"';
access_log /var/log/nginx/access.log main;
error_log /var/log/nginx/error.log;
# 基础配置
sendfile on;
tcp_nopush on;
tcp_nodelay on;
keepalive_timeout 65;
types_hash_max_size 2048;
# Gzip压缩
gzip on;
gzip_vary on;
gzip_min_length 1024;
gzip_proxied any;
gzip_comp_level 6;
gzip_types
text/plain
text/css
text/xml
text/javascript
application/json
application/javascript
application/xml+rss
application/atom+xml
image/svg+xml;
# 上游服务器配置
upstream ruoyi_backend {
# 轮询负载均衡
server ruoyi-admin-9900:9900 weight=1 max_fails=3 fail_timeout=30s;
server ruoyi-admin-9901:9901 weight=1 max_fails=3 fail_timeout=30s;
server ruoyi-admin-9902:9902 weight=1 max_fails=3 fail_timeout=30s;
server ruoyi-admin-9903:9903 weight=1 max_fails=3 fail_timeout=30s;
# 健康检查
keepalive 32;
}
# 主服务器配置
server {
listen 80;
server_name localhost;
# 静态资源缓存
location ~* \.(js|css|png|jpg|jpeg|gif|ico|svg)$ {
expires 1y;
add_header Cache-Control "public, immutable";
proxy_pass http://ruoyi_backend;
}
# API请求
location / {
proxy_pass http://ruoyi_backend;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
# 超时设置
proxy_connect_timeout 30s;
proxy_send_timeout 30s;
proxy_read_timeout 30s;
# 缓冲设置
proxy_buffering on;
proxy_buffer_size 4k;
proxy_buffers 8 4k;
proxy_busy_buffers_size 8k;
}
# 健康检查端点
location /health {
access_log off;
return 200 "healthy\n";
add_header Content-Type text/plain;
}
# 状态监控页面
location /nginx_status {
stub_status on;
access_log off;
allow 127.0.0.1;
deny all;
}
}
# 监控和日志配置
server {
listen 8080;
server_name localhost;
location / {
return 200 "Nginx Load Balancer Status\n";
add_header Content-Type text/plain;
}
location /status {
stub_status on;
access_log off;
}
}
}
@@ -0,0 +1,368 @@
#!/bin/bash
# 扩展服务启动停止一体化脚本
# 使用方法: ./start-extend-services.sh [start|stop|restart|status] [service_name]
# 服务名称: monitor, snailjob, all
# 配置变量
MONITOR_JAR="ruoyi-monitor-admin.jar"
SNAILJOB_JAR="ruoyi-snailjob-server.jar"
JAVA_OPTS="-Xms2048m -Xmx2048m -XX:MetaspaceSize=512m -XX:MaxMetaspaceSize=1024m"
JAVA_OPTS="$JAVA_OPTS -Dfile.encoding=UTF-8 -Duser.timezone=Asia/Shanghai"
JAVA_OPTS="$JAVA_OPTS -XX:+UseG1GC -XX:MaxGCPauseMillis=200"
JAVA_OPTS="$JAVA_OPTS -XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=logs/"
# 服务配置
declare -A SERVICES
SERVICES[monitor]="9090:ruoyi-monitor-admin:prod"
SERVICES[snailjob]="8800:ruoyi-snailjob-server:prod"
# 日志目录
LOG_DIR="logs"
mkdir -p $LOG_DIR
# 颜色输出
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
NC='\033[0m' # No Color
# 打印带颜色的消息
print_message() {
local color=$1
local message=$2
echo -e "${color}${message}${NC}"
}
# 检查JAR文件是否存在
check_jar() {
local service=$1
local jar_file=""
case $service in
monitor)
jar_file="./$MONITOR_JAR"
;;
snailjob)
jar_file="./$SNAILJOB_JAR"
;;
esac
if [ ! -f "$jar_file" ]; then
print_message $RED "错误: $jar_file 文件不存在!"
print_message $YELLOW "请先编译项目: mvn clean package -Pprod"
return 1
fi
return 0
}
# 启动单个服务
start_service() {
local service=$1
local config=${SERVICES[$service]}
local port=$(echo $config | cut -d: -f1)
local app_name=$(echo $config | cut -d: -f2)
local profile=$(echo $config | cut -d: -f3)
local pid_file="logs/${app_name}.pid"
local log_file="logs/${app_name}.log"
# 检查是否已经运行
if [ -f "$pid_file" ]; then
local pid=$(cat "$pid_file")
if ps -p $pid > /dev/null 2>&1; then
print_message $YELLOW "服务 $service 已经在运行中 (PID: $pid)"
return
else
# PID文件存在但进程不存在,清理PID文件
rm -f "$pid_file"
fi
fi
# 检查JAR文件
if ! check_jar $service; then
return 1
fi
print_message $GREEN "正在启动服务 $service (端口: $port, 配置文件: $profile)..."
# 启动应用
case $service in
monitor)
local jar_path="../ruoyi-extend/ruoyi-monitor-admin/target/$MONITOR_JAR"
nohup java $JAVA_OPTS \
-Dspring.profiles.active=$profile \
-Dserver.port=$port \
-jar "$jar_path" \
> "$log_file" 2>&1 &
;;
snailjob)
local jar_path="../ruoyi-extend/ruoyi-snailjob-server/target/$SNAILJOB_JAR"
nohup java $JAVA_OPTS \
-Dspring.profiles.active=$profile \
-Dserver.port=$port \
-jar "$jar_path" \
> "$log_file" 2>&1 &
;;
esac
local new_pid=$!
echo $new_pid > "$pid_file"
# 等待启动
sleep 5
# 检查启动状态
if ps -p $new_pid > /dev/null 2>&1; then
print_message $GREEN "服务 $service 启动成功 (PID: $new_pid)"
print_message $GREEN "访问地址: http://localhost:$port"
print_message $GREEN "日志文件: $log_file"
# 显示服务特定信息
case $service in
monitor)
print_message $BLUE "监控中心: http://localhost:$port/admin"
;;
snailjob)
print_message $BLUE "任务调度: http://localhost:$port/snail-job"
;;
esac
else
print_message $RED "服务 $service 启动失败,请检查日志文件: $log_file"
rm -f "$pid_file"
fi
}
# 停止单个服务
stop_service() {
local service=$1
local config=${SERVICES[$service]}
local app_name=$(echo $config | cut -d: -f2)
local pid_file="logs/${app_name}.pid"
if [ -f "$pid_file" ]; then
local pid=$(cat "$pid_file")
if ps -p $pid > /dev/null 2>&1; then
print_message $YELLOW "正在停止服务 $service (PID: $pid)..."
kill $pid
# 等待进程结束
local count=0
while ps -p $pid > /dev/null 2>&1 && [ $count -lt 30 ]; do
sleep 1
count=$((count + 1))
done
# 强制杀死进程
if ps -p $pid > /dev/null 2>&1; then
print_message $YELLOW "强制停止服务 $service..."
kill -9 $pid
fi
rm -f "$pid_file"
print_message $GREEN "服务 $service 已停止"
else
print_message $YELLOW "服务 $service 未运行"
rm -f "$pid_file"
fi
else
print_message $YELLOW "服务 $service 未运行"
fi
}
# 重启单个服务
restart_service() {
local service=$1
print_message $YELLOW "重启服务 $service..."
stop_service $service
sleep 3
start_service $service
}
# 查看服务状态
status_service() {
local service=$1
local config=${SERVICES[$service]}
local port=$(echo $config | cut -d: -f1)
local app_name=$(echo $config | cut -d: -f2)
local pid_file="logs/${app_name}.pid"
if [ -f "$pid_file" ]; then
local pid=$(cat "$pid_file")
if ps -p $pid > /dev/null 2>&1; then
local memory=$(ps -o rss= -p $pid | awk '{print $1/1024 " MB"}')
local uptime=$(ps -o etime= -p $pid)
print_message $GREEN "服务 $service: 运行中 (PID: $pid, 端口: $port, 内存: $memory, 运行时间: $uptime)"
else
print_message $RED "服务 $service: 异常 (PID文件存在但进程不存在)"
rm -f "$pid_file"
fi
else
print_message $YELLOW "服务 $service: 未运行 (端口: $port)"
fi
}
# 启动所有服务
start_all() {
print_message $GREEN "启动所有扩展服务..."
for service in "${!SERVICES[@]}"; do
start_service $service
sleep 3
done
}
# 停止所有服务
stop_all() {
print_message $YELLOW "停止所有扩展服务..."
for service in "${!SERVICES[@]}"; do
stop_service $service
done
}
# 重启所有服务
restart_all() {
print_message $YELLOW "重启所有扩展服务..."
stop_all
sleep 5
start_all
}
# 查看所有服务状态
status_all() {
print_message $GREEN "所有扩展服务状态:"
for service in "${!SERVICES[@]}"; do
status_service $service
done
}
# 编译项目
build_project() {
print_message $BLUE "开始编译扩展服务项目..."
# 编译monitor-admin
print_message $YELLOW "编译 ruoyi-monitor-admin..."
cd ../ruoyi-extend/ruoyi-monitor-admin
mvn clean package -Pprod
if [ $? -eq 0 ]; then
print_message $GREEN "ruoyi-monitor-admin 编译成功"
else
print_message $RED "ruoyi-monitor-admin 编译失败"
return 1
fi
# 编译snailjob-server
print_message $YELLOW "编译 ruoyi-snailjob-server..."
cd ../ruoyi-snailjob-server
mvn clean package -Pprod
if [ $? -eq 0 ]; then
print_message $GREEN "ruoyi-snailjob-server 编译成功"
else
print_message $RED "ruoyi-snailjob-server 编译失败"
return 1
fi
# 返回原目录
cd ../../部署文档/杭州第一技师部署文档
print_message $GREEN "所有扩展服务编译完成"
}
# 显示帮助信息
show_help() {
echo "扩展服务启动停止一体化脚本使用方法:"
echo " $0 [命令] [服务名称]"
echo ""
echo "命令:"
echo " start 启动服务"
echo " stop 停止服务"
echo " restart 重启服务"
echo " status 查看服务状态"
echo " build 编译项目"
echo ""
echo "服务名称:"
echo " monitor 启动监控中心服务 (端口: 9090)"
echo " snailjob 启动任务调度服务 (端口: 8800)"
echo " all 操作所有服务"
echo ""
echo "示例:"
echo " $0 start monitor # 启动监控中心服务"
echo " $0 start snailjob # 启动任务调度服务"
echo " $0 start all # 启动所有服务"
echo " $0 stop all # 停止所有服务"
echo " $0 status all # 查看所有服务状态"
echo " $0 build # 编译所有扩展服务"
echo ""
echo "服务访问地址:"
echo " 监控中心: http://localhost:9090/admin"
echo " 任务调度: http://localhost:8800/snail-job"
}
# 主函数
main() {
local action=$1
local service=$2
# 检查参数
if [ -z "$action" ]; then
show_help
exit 1
fi
case $action in
start)
if [ "$service" = "all" ]; then
start_all
elif [ -n "${SERVICES[$service]}" ]; then
start_service $service
else
print_message $RED "无效的服务名称: $service"
show_help
exit 1
fi
;;
stop)
if [ "$service" = "all" ]; then
stop_all
elif [ -n "${SERVICES[$service]}" ]; then
stop_service $service
else
print_message $RED "无效的服务名称: $service"
show_help
exit 1
fi
;;
restart)
if [ "$service" = "all" ]; then
restart_all
elif [ -n "${SERVICES[$service]}" ]; then
restart_service $service
else
print_message $RED "无效的服务名称: $service"
show_help
exit 1
fi
;;
status)
if [ "$service" = "all" ]; then
status_all
elif [ -n "${SERVICES[$service]}" ]; then
status_service $service
else
print_message $RED "无效的服务名称: $service"
show_help
exit 1
fi
;;
build)
build_project
;;
*)
print_message $RED "无效的命令: $action"
show_help
exit 1
;;
esac
}
# 执行主函数
main "$@"
@@ -0,0 +1,296 @@
@echo off
chcp 65001 >nul
setlocal enabledelayedexpansion
REM 多实例启动脚本 (Windows版本)
REM 使用方法: start-multi-instance.bat [start|stop|restart|status] [instance_name]
REM 实例名称: 9900, 9901, 9902, 9903, all
REM 配置变量
set JAR_NAME=ruoyi-admin.jar
set JAVA_OPTS=-Xms512m -Xmx1024m -XX:MetaspaceSize=128m -XX:MaxMetaspaceSize=256m
set JAVA_OPTS=%JAVA_OPTS% -Dfile.encoding=UTF-8 -Duser.timezone=Asia/Shanghai
set JAVA_OPTS=%JAVA_OPTS% -XX:+UseG1GC -XX:MaxGCPauseMillis=200
set JAVA_OPTS=%JAVA_OPTS% -XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=logs\
REM 日志目录
if not exist logs mkdir logs
REM 颜色输出函数
:print_message
set color=%1
set message=%2
if "%color%"=="RED" (
echo [31m%message%[0m
) else if "%color%"=="GREEN" (
echo [32m%message%[0m
) else if "%color%"=="YELLOW" (
echo [33m%message%[0m
) else (
echo %message%
)
goto :eof
REM 检查JAR文件是否存在
:check_jar
if not exist "%JAR_NAME%" (
call :print_message RED "错误: %JAR_NAME% 文件不存在!"
call :print_message YELLOW "请确保在正确的目录下运行此脚本,或先执行 mvn clean package -Pprod 编译项目"
pause
exit /b 1
)
goto :eof
REM 启动单个实例
:start_instance
set port=%1
set profile=%2
set pid_file=logs\ruoyi-admin-%port%.pid
set log_file=logs\ruoyi-admin-%port%.log
REM 检查是否已经运行
if exist "%pid_file%" (
for /f "tokens=*" %%i in (%pid_file%) do set pid=%%i
tasklist /FI "PID eq !pid!" 2>nul | find /I "!pid!" >nul
if !errorlevel! equ 0 (
call :print_message YELLOW "实例 %port% 已经在运行中 (PID: !pid!)"
goto :eof
) else (
REM PID文件存在但进程不存在,清理PID文件
del "%pid_file%" 2>nul
)
)
call :print_message GREEN "正在启动实例 %port% (配置文件: %profile%)..."
REM 启动应用
start /b java %JAVA_OPTS% -Dspring.profiles.active=%profile% -Dserver.port=%port% -jar %JAR_NAME% > "%log_file%" 2>&1
REM 等待启动
timeout /t 3 /nobreak >nul
REM 查找新启动的进程
for /f "tokens=2" %%i in ('tasklist /FI "IMAGENAME eq java.exe" /FO CSV ^| findstr /I "java.exe"') do (
set new_pid=%%i
set new_pid=!new_pid:"=!
echo !new_pid! > "%pid_file%"
call :print_message GREEN "实例 %port% 启动成功 (PID: !new_pid!)"
call :print_message GREEN "访问地址: http://localhost:%port%"
call :print_message GREEN "日志文件: %log_file%"
goto :eof
)
call :print_message RED "实例 %port% 启动失败,请检查日志文件: %log_file%"
if exist "%pid_file%" del "%pid_file%"
goto :eof
REM 停止单个实例
:stop_instance
set port=%1
set pid_file=logs\ruoyi-admin-%port%.pid
if exist "%pid_file%" (
for /f "tokens=*" %%i in (%pid_file%) do set pid=%%i
tasklist /FI "PID eq !pid!" 2>nul | find /I "!pid!" >nul
if !errorlevel! equ 0 (
call :print_message YELLOW "正在停止实例 %port% (PID: !pid!)..."
taskkill /PID !pid! /F >nul 2>&1
del "%pid_file%" 2>nul
call :print_message GREEN "实例 %port% 已停止"
) else (
call :print_message YELLOW "实例 %port% 未运行"
del "%pid_file%" 2>nul
)
) else (
call :print_message YELLOW "实例 %port% 未运行"
)
goto :eof
REM 重启单个实例
:restart_instance
set port=%1
call :print_message YELLOW "重启实例 %port%..."
call :stop_instance %port%
timeout /t 2 /nobreak >nul
if "%port%"=="9900" call :start_instance %port% prod-9900
if "%port%"=="9901" call :start_instance %port% prod-9901
if "%port%"=="9902" call :start_instance %port% prod-9902
if "%port%"=="9903" call :start_instance %port% prod-9903
goto :eof
REM 查看实例状态
:status_instance
set port=%1
set pid_file=logs\ruoyi-admin-%port%.pid
if exist "%pid_file%" (
for /f "tokens=*" %%i in (%pid_file%) do set pid=%%i
tasklist /FI "PID eq !pid!" 2>nul | find /I "!pid!" >nul
if !errorlevel! equ 0 (
call :print_message GREEN "实例 %port%: 运行中 (PID: !pid!)"
) else (
call :print_message RED "实例 %port%: 异常 (PID文件存在但进程不存在)"
del "%pid_file%" 2>nul
)
) else (
call :print_message YELLOW "实例 %port%: 未运行"
)
goto :eof
REM 启动所有实例
:start_all
call :print_message GREEN "启动所有实例..."
call :start_instance 9900 prod-9900
timeout /t 2 /nobreak >nul
call :start_instance 9901 prod-9901
timeout /t 2 /nobreak >nul
call :start_instance 9902 prod-9902
timeout /t 2 /nobreak >nul
call :start_instance 9903 prod-9903
goto :eof
REM 停止所有实例
:stop_all
call :print_message YELLOW "停止所有实例..."
call :stop_instance 9900
call :stop_instance 9901
call :stop_instance 9902
call :stop_instance 9903
goto :eof
REM 重启所有实例
:restart_all
call :print_message YELLOW "重启所有实例..."
call :stop_all
timeout /t 3 /nobreak >nul
call :start_all
goto :eof
REM 查看所有实例状态
:status_all
call :print_message GREEN "所有实例状态:"
call :status_instance 9900
call :status_instance 9901
call :status_instance 9902
call :status_instance 9903
goto :eof
REM 显示帮助信息
:show_help
echo 多实例启动脚本使用方法:
echo %0 [命令] [实例名称]
echo.
echo 命令:
echo start 启动实例
echo stop 停止实例
echo restart 重启实例
echo status 查看实例状态
echo.
echo 实例名称:
echo 9900 启动9900端口实例
echo 9901 启动9901端口实例
echo 9902 启动9902端口实例
echo 9903 启动9903端口实例
echo all 操作所有实例
echo.
echo 示例:
echo %0 start 9900 # 启动9900端口实例
echo %0 start all # 启动所有实例
echo %0 stop 9900 # 停止9900端口实例
echo %0 status all # 查看所有实例状态
echo %0 restart all # 重启所有实例
goto :eof
REM 主函数
:main
set action=%1
set instance=%2
REM 检查参数
if "%action%"=="" (
call :show_help
pause
exit /b 1
)
REM 检查JAR文件
call :check_jar
if !errorlevel! neq 0 goto :eof
if "%action%"=="start" (
if "%instance%"=="all" (
call :start_all
) else if "%instance%"=="9900" (
call :start_instance 9900 prod-9900
) else if "%instance%"=="9901" (
call :start_instance 9901 prod-9901
) else if "%instance%"=="9902" (
call :start_instance 9902 prod-9902
) else if "%instance%"=="9903" (
call :start_instance 9903 prod-9903
) else (
call :print_message RED "无效的实例名称: %instance%"
call :show_help
pause
exit /b 1
)
) else if "%action%"=="stop" (
if "%instance%"=="all" (
call :stop_all
) else if "%instance%"=="9900" (
call :stop_instance 9900
) else if "%instance%"=="9901" (
call :stop_instance 9901
) else if "%instance%"=="9902" (
call :stop_instance 9902
) else if "%instance%"=="9903" (
call :stop_instance 9903
) else (
call :print_message RED "无效的实例名称: %instance%"
call :show_help
pause
exit /b 1
)
) else if "%action%"=="restart" (
if "%instance%"=="all" (
call :restart_all
) else if "%instance%"=="9900" (
call :restart_instance 9900
) else if "%instance%"=="9901" (
call :restart_instance 9901
) else if "%instance%"=="9902" (
call :restart_instance 9902
) else if "%instance%"=="9903" (
call :restart_instance 9903
) else (
call :print_message RED "无效的实例名称: %instance%"
call :show_help
pause
exit /b 1
)
) else if "%action%"=="status" (
if "%instance%"=="all" (
call :status_all
) else if "%instance%"=="9900" (
call :status_instance 9900
) else if "%instance%"=="9901" (
call :status_instance 9901
) else if "%instance%"=="9902" (
call :status_instance 9902
) else if "%instance%"=="9903" (
call :status_instance 9903
) else (
call :print_message RED "无效的实例名称: %instance%"
call :show_help
pause
exit /b 1
)
) else (
call :print_message RED "无效的命令: %action%"
call :show_help
pause
exit /b 1
)
pause
goto :eof
@@ -0,0 +1,284 @@
#!/bin/bash
# 多实例启动脚本
# 使用方法: ./start-multi-instance.sh [start|stop|restart|status] [instance_name]
# 实例名称: 9900, 9901, 9902, 9903, all
# 配置变量
JAR_NAME="ruoyi-admin.jar"
JAVA_OPTS="-Xms8192m -Xmx8192m -XX:MetaspaceSize=512m -XX:MaxMetaspaceSize=1024m"
JAVA_OPTS="$JAVA_OPTS -Dfile.encoding=UTF-8 -Duser.timezone=Asia/Shanghai"
JAVA_OPTS="$JAVA_OPTS -XX:+UseG1GC -XX:MaxGCPauseMillis=200"
JAVA_OPTS="$JAVA_OPTS -XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=logs/"
# 实例配置
declare -A INSTANCES
INSTANCES[9900]="prod-9900"
INSTANCES[9901]="prod-9901"
INSTANCES[9902]="prod-9902"
INSTANCES[9903]="prod-9903"
# 日志目录
LOG_DIR="logs"
mkdir -p $LOG_DIR
# 颜色输出
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
NC='\033[0m' # No Color
# 打印带颜色的消息
print_message() {
local color=$1
local message=$2
echo -e "${color}${message}${NC}"
}
# 检查JAR文件是否存在
check_jar() {
if [ ! -f "$JAR_NAME" ]; then
print_message $RED "错误: $JAR_NAME 文件不存在!"
print_message $YELLOW "请确保在正确的目录下运行此脚本,或先执行 mvn clean package -Pprod 编译项目"
exit 1
fi
}
# 启动单个实例
start_instance() {
local port=$1
local profile=$2
local pid_file="logs/ruoyi-admin-${port}.pid"
local log_file="logs/ruoyi-admin-${port}.log"
# 检查是否已经运行
if [ -f "$pid_file" ]; then
local pid=$(cat "$pid_file")
if ps -p $pid > /dev/null 2>&1; then
print_message $YELLOW "实例 $port 已经在运行中 (PID: $pid)"
return
else
# PID文件存在但进程不存在,清理PID文件
rm -f "$pid_file"
fi
fi
print_message $GREEN "正在启动实例 $port (配置文件: $profile)..."
# 启动应用
nohup java $JAVA_OPTS \
-Dspring.profiles.active=prod,$profile \
-Dserver.port=$port \
-jar $JAR_NAME \
> "$log_file" 2>&1 &
local new_pid=$!
echo $new_pid > "$pid_file"
# 等待启动
sleep 3
# 检查启动状态
if ps -p $new_pid > /dev/null 2>&1; then
print_message $GREEN "实例 $port 启动成功 (PID: $new_pid)"
print_message $GREEN "访问地址: http://localhost:$port"
print_message $GREEN "日志文件: $log_file"
else
print_message $RED "实例 $port 启动失败,请检查日志文件: $log_file"
rm -f "$pid_file"
fi
}
# 停止单个实例
stop_instance() {
local port=$1
local pid_file="logs/ruoyi-admin-${port}.pid"
if [ -f "$pid_file" ]; then
local pid=$(cat "$pid_file")
if ps -p $pid > /dev/null 2>&1; then
print_message $YELLOW "正在停止实例 $port (PID: $pid)..."
kill $pid
# 等待进程结束
local count=0
while ps -p $pid > /dev/null 2>&1 && [ $count -lt 30 ]; do
sleep 1
count=$((count + 1))
done
# 强制杀死进程
if ps -p $pid > /dev/null 2>&1; then
print_message $YELLOW "强制停止实例 $port..."
kill -9 $pid
fi
rm -f "$pid_file"
print_message $GREEN "实例 $port 已停止"
else
print_message $YELLOW "实例 $port 未运行"
rm -f "$pid_file"
fi
else
print_message $YELLOW "实例 $port 未运行"
fi
}
# 重启单个实例
restart_instance() {
local port=$1
print_message $YELLOW "重启实例 $port..."
stop_instance $port
sleep 2
start_instance $port ${INSTANCES[$port]}
}
# 查看实例状态
status_instance() {
local port=$1
local pid_file="logs/ruoyi-admin-${port}.pid"
if [ -f "$pid_file" ]; then
local pid=$(cat "$pid_file")
if ps -p $pid > /dev/null 2>&1; then
local memory=$(ps -o rss= -p $pid | awk '{print $1/1024 " MB"}')
local uptime=$(ps -o etime= -p $pid)
print_message $GREEN "实例 $port: 运行中 (PID: $pid, 内存: $memory, 运行时间: $uptime)"
else
print_message $RED "实例 $port: 异常 (PID文件存在但进程不存在)"
rm -f "$pid_file"
fi
else
print_message $YELLOW "实例 $port: 未运行"
fi
}
# 启动所有实例
start_all() {
print_message $GREEN "启动所有实例..."
for port in "${!INSTANCES[@]}"; do
start_instance $port ${INSTANCES[$port]}
sleep 2
done
}
# 停止所有实例
stop_all() {
print_message $YELLOW "停止所有实例..."
for port in "${!INSTANCES[@]}"; do
stop_instance $port
done
}
# 重启所有实例
restart_all() {
print_message $YELLOW "重启所有实例..."
stop_all
sleep 3
start_all
}
# 查看所有实例状态
status_all() {
print_message $GREEN "所有实例状态:"
for port in "${!INSTANCES[@]}"; do
status_instance $port
done
}
# 显示帮助信息
show_help() {
echo "多实例启动脚本使用方法:"
echo " $0 [命令] [实例名称]"
echo ""
echo "命令:"
echo " start 启动实例"
echo " stop 停止实例"
echo " restart 重启实例"
echo " status 查看实例状态"
echo ""
echo "实例名称:"
echo " 9900 启动9900端口实例"
echo " 9901 启动9901端口实例"
echo " 9902 启动9902端口实例"
echo " 9903 启动9903端口实例"
echo " all 操作所有实例"
echo ""
echo "示例:"
echo " $0 start 9900 # 启动9900端口实例"
echo " $0 start all # 启动所有实例"
echo " $0 stop 9900 # 停止9900端口实例"
echo " $0 status all # 查看所有实例状态"
echo " $0 restart all # 重启所有实例"
}
# 主函数
main() {
local action=$1
local instance=$2
# 检查参数
if [ -z "$action" ]; then
show_help
exit 1
fi
# 检查JAR文件
check_jar
case $action in
start)
if [ "$instance" = "all" ]; then
start_all
elif [ -n "${INSTANCES[$instance]}" ]; then
start_instance $instance ${INSTANCES[$instance]}
else
print_message $RED "无效的实例名称: $instance"
show_help
exit 1
fi
;;
stop)
if [ "$instance" = "all" ]; then
stop_all
elif [ -n "${INSTANCES[$instance]}" ]; then
stop_instance $instance
else
print_message $RED "无效的实例名称: $instance"
show_help
exit 1
fi
;;
restart)
if [ "$instance" = "all" ]; then
restart_all
elif [ -n "${INSTANCES[$instance]}" ]; then
restart_instance $instance
else
print_message $RED "无效的实例名称: $instance"
show_help
exit 1
fi
;;
status)
if [ "$instance" = "all" ]; then
status_all
elif [ -n "${INSTANCES[$instance]}" ]; then
status_instance $instance
else
print_message $RED "无效的实例名称: $instance"
show_help
exit 1
fi
;;
*)
print_message $RED "无效的命令: $action"
show_help
exit 1
;;
esac
}
# 执行主函数
main "$@"
@@ -0,0 +1,66 @@
#!/bin/bash
# 配置修复验证脚本
# 使用方法: ./test-config-fix.sh
# 颜色输出
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
NC='\033[0m' # No Color
print_message() {
local color=$1
local message=$2
echo -e "${color}${message}${NC}"
}
print_message $GREEN "=== 多实例配置修复验证脚本 ==="
# 检查JAR文件
JAR_NAME="ruoyi-admin.jar"
if [ ! -f "$JAR_NAME" ]; then
print_message $RED "错误: $JAR_NAME 文件不存在!"
print_message $YELLOW "请先执行 mvn clean package -Pprod 编译项目"
exit 1
fi
print_message $GREEN "✓ JAR文件存在"
# 检查配置文件
CONFIG_FILES=(
"backend/ruoyi-admin/src/main/resources/application-prod-9900.yml"
"backend/ruoyi-admin/src/main/resources/application-prod-9901.yml"
"backend/ruoyi-admin/src/main/resources/application-prod-9902.yml"
"backend/ruoyi-admin/src/main/resources/application-prod-9903.yml"
)
for config_file in "${CONFIG_FILES[@]}"; do
if [ -f "$config_file" ]; then
# 检查是否包含spring.profiles.include
if grep -q "spring.profiles.include" "$config_file"; then
print_message $RED "✗ $config_file 仍包含 spring.profiles.include"
else
print_message $GREEN "✓ $config_file 配置正确"
fi
else
print_message $RED "✗ $config_file 文件不存在"
fi
done
# 检查启动脚本
START_SCRIPT="docs/部署文档/杭州第一技师部署文档/start-multi-instance.sh"
if [ -f "$START_SCRIPT" ]; then
if grep -q "spring.profiles.active=prod,\$profile" "$START_SCRIPT"; then
print_message $GREEN "✓ 启动脚本配置正确"
else
print_message $RED "✗ 启动脚本配置错误"
fi
else
print_message $RED "✗ 启动脚本不存在"
fi
print_message $GREEN ""
print_message $GREEN "=== 验证完成 ==="
print_message $YELLOW "如果所有检查都通过,可以尝试启动实例:"
print_message $YELLOW " ./start-multi-instance.sh start 9900"
@@ -0,0 +1,96 @@
# 多实例启动配置修复说明
## 问题描述
在使用 `start-multi-instance.sh` 启动实例时,出现以下错误:
```
org.springframework.boot.context.config.InvalidConfigDataPropertyException: Property 'spring.profiles.include' imported from location 'class path resource [application-prod-9900.yml]' is invalid in a profile specific resource
```
## 问题原因
在Spring Boot 2.4+版本中,`spring.profiles.include` 属性不能在profile-specific的配置文件中使用(如 `application-prod-9900.yml`)。这个属性只能在主配置文件(`application.yml`)中使用。
## 修复方案
### 1. 修改配置文件结构
移除了所有实例配置文件中的 `spring.profiles.include: prod` 配置:
**修复前:**
```yaml
# application-prod-9900.yml
spring:
profiles:
include: prod # ❌ 这在profile-specific文件中不被允许
```
**修复后:**
```yaml
# application-prod-9900.yml
spring:
application:
name: RuoYi-Vue-Plus-9900 # ✅ 直接配置应用名称
```
### 2. 修改启动脚本
修改启动脚本,在启动时同时激活 `prod` 和具体的实例profile:
**修复前:**
```bash
-Dspring.profiles.active=$profile # 只激活 prod-9900
```
**修复后:**
```bash
-Dspring.profiles.active=prod,$profile # 同时激活 prod 和 prod-9900
```
## 修复的文件
1. `backend/ruoyi-admin/src/main/resources/application-prod-9900.yml`
2. `backend/ruoyi-admin/src/main/resources/application-prod-9901.yml`
3. `backend/ruoyi-admin/src/main/resources/application-prod-9902.yml`
4. `backend/ruoyi-admin/src/main/resources/application-prod-9903.yml`
5. `docs/部署文档/杭州第一技师部署文档/start-multi-instance.sh`
## 配置加载顺序
修复后的配置加载顺序:
1. `application.yml` (主配置文件)
2. `application-prod.yml` (生产环境基础配置)
3. `application-prod-9900.yml` (9900端口实例特定配置)
## 验证方法
1. 重新编译项目:
```bash
mvn clean package -Pprod
```
2. 启动单个实例测试:
```bash
./start-multi-instance.sh start 9900
```
3. 检查启动日志,确认没有配置错误
4. 访问应用验证功能正常:
```bash
curl http://localhost:9900
```
## 注意事项
1. 确保在修改配置文件后重新编译项目
2. 如果之前有实例在运行,需要先停止再重新启动
3. 检查日志文件确认配置加载正确
4. 验证监控中心和SnailJob配置是否正常工作
## 相关文档
- [Spring Boot Profile配置最佳实践](https://docs.spring.io/spring-boot/docs/current/reference/html/features.html#features.external-config.files.profile-specific)
- [多实例部署指南](../README.md)
@@ -0,0 +1,293 @@
# 🚀 RuoYi-Admin 多实例部署说明
## 📋 概述
本文档说明如何在生产环境中部署多个 `ruoyi-admin.jar` 实例到不同端口,实现负载均衡和高可用。
## 🏗️ 部署架构
```
┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
│ 前端应用 │ │ Nginx │ │ 后端实例 │
│ │ │ 负载均衡器 │ │ │
│ http://80 │───▶│ http://80 │───▶│ 9900, 9901 │
│ │ │ │ │ 9902, 9903 │
└─────────────────┘ └─────────────────┘ └─────────────────┘
```
## 📁 文件结构
```
backend/
├── ruoyi-admin.jar # 编译后的JAR包
├── application-prod-9900.yml # 9900端口配置文件
├── application-prod-9901.yml # 9901端口配置文件
├── application-prod-9902.yml # 9902端口配置文件
├── application-prod-9903.yml # 9903端口配置文件
├── start-multi-instance.sh # Linux启动脚本
├── start-multi-instance.bat # Windows启动脚本
├── docker-compose-multi.yml # Docker部署配置
├── nginx.conf # Nginx负载均衡配置
└── logs/ # 日志目录
├── ruoyi-admin-9900.log
├── ruoyi-admin-9901.log
├── ruoyi-admin-9902.log
└── ruoyi-admin-9903.log
```
## 🔧 部署步骤
### 1. 编译项目
```bash
# 进入backend目录
cd backend
# 编译生产环境JAR包
mvn clean package -Pprod
# 检查生成的JAR文件
ls -la ruoyi-admin/target/ruoyi-admin.jar
```
### 2. 准备部署文件
```bash
# 复制JAR文件到部署目录
cp ruoyi-admin/target/ruoyi-admin.jar ./
# 创建日志目录
mkdir -p logs
# 设置脚本执行权限 (Linux)
chmod +x start-multi-instance.sh
```
### 3. 启动多实例
#### Linux环境
```bash
# 启动所有实例
./start-multi-instance.sh start all
# 启动单个实例
./start-multi-instance.sh start 9900
./start-multi-instance.sh start 9901
./start-multi-instance.sh start 9902
./start-multi-instance.sh start 9903
# 查看实例状态
./start-multi-instance.sh status all
# 停止实例
./start-multi-instance.sh stop all
# 重启实例
./start-multi-instance.sh restart all
```
#### Windows环境
```cmd
# 启动所有实例
start-multi-instance.bat start all
# 启动单个实例
start-multi-instance.bat start 9900
start-multi-instance.bat start 9901
start-multi-instance.bat start 9902
start-multi-instance.bat start 9903
# 查看实例状态
start-multi-instance.bat status all
# 停止实例
start-multi-instance.bat stop all
# 重启实例
start-multi-instance.bat restart all
```
### 4. Docker部署 (可选)
```bash
# 启动所有服务
docker-compose -f docker-compose-multi.yml up -d
# 查看服务状态
docker-compose -f docker-compose-multi.yml ps
# 查看日志
docker-compose -f docker-compose-multi.yml logs -f
# 停止服务
docker-compose -f docker-compose-multi.yml down
```
## 🌐 访问地址
### 直接访问各实例
- **实例1**: http://localhost:9900
- **实例2**: http://localhost:9901
- **实例3**: http://localhost:9902
- **实例4**: http://localhost:9903
### 通过负载均衡器访问
- **主入口**: http://localhost (Nginx负载均衡)
- **监控页面**: http://localhost:8080/status
## 📊 监控和管理
### 1. 实例状态监控
```bash
# 查看所有实例状态
./start-multi-instance.sh status all
# 查看单个实例状态
./start-multi-instance.sh status 9900
```
### 2. 日志监控
```bash
# 查看实时日志
tail -f logs/ruoyi-admin-9900.log
tail -f logs/ruoyi-admin-9901.log
tail -f logs/ruoyi-admin-9902.log
tail -f logs/ruoyi-admin-9903.log
# 查看错误日志
grep "ERROR" logs/ruoyi-admin-*.log
```
### 3. 进程监控
```bash
# 查看Java进程
ps aux | grep java
# 查看端口占用
netstat -tlnp | grep :99
```
## ⚠️ 注意事项
### 1. 端口冲突
- 确保9900-9903端口未被其他服务占用
- 检查防火墙设置,确保端口可访问
### 2. 内存配置
- 每个实例默认配置:最小512MB,最大1024MB
- 根据服务器配置调整 `JAVA_OPTS` 参数
### 3. 数据库连接
- 所有实例共享同一个数据库
- 确保数据库连接池配置合理,避免连接数过多
### 4. 文件上传
- 临时文件目录:`/ruoyi/server/temp`
- 确保目录存在且有写权限
## 🔄 故障排查
### 1. 实例启动失败
```bash
# 检查JAR文件是否存在
ls -la ruoyi-admin.jar
# 检查端口是否被占用
netstat -tlnp | grep :99
# 查看启动日志
tail -f logs/ruoyi-admin-9900.log
```
### 2. 实例无法访问
```bash
# 检查防火墙设置
firewall-cmd --list-ports
# 检查SELinux状态 (CentOS/RHEL)
getenforce
# 测试端口连通性
telnet localhost 9900
```
### 3. 负载均衡不工作
```bash
# 检查Nginx状态
docker-compose -f docker-compose-multi.yml ps nginx
# 查看Nginx日志
docker-compose -f docker-compose-multi.yml logs nginx
# 测试后端服务
curl http://localhost:9900/health
```
## 📈 性能优化建议
### 1. JVM调优
```bash
# 根据服务器内存调整JVM参数
JAVA_OPTS="-Xms2g -Xmx4g -XX:MetaspaceSize=256m -XX:MaxMetaspaceSize=512m"
```
### 2. 数据库连接池优化
```yaml
# 在application-prod-*.yml中调整
spring:
datasource:
hikari:
maxPoolSize: 100 # 根据实例数调整
minIdle: 20
connectionTimeout: 30000
```
### 3. 缓存配置
```yaml
# Redis连接池优化
redisson:
threads: 32
nettyThreads: 64
singleServerConfig:
connectionPoolSize: 128
```
## 🚀 扩展部署
### 1. 添加更多实例
- 复制配置文件模板
- 修改端口号和应用名称
- 更新启动脚本和Docker配置
### 2. 集群部署
- 使用Redis集群
- 配置数据库主从复制
- 实现会话共享
### 3. 容器化部署
- 使用Kubernetes管理
- 配置自动扩缩容
- 实现滚动更新
## 📞 技术支持
如遇到部署问题,请检查:
1. 系统日志和应用日志
2. 端口占用情况
3. 防火墙和网络配置
4. 数据库连接状态
5. 内存和磁盘空间
---
**📝 文档版本**: v1.0
**🔄 最后更新**: 2025-01-03
**👨‍💻 维护人员**: 开发团队
@@ -0,0 +1,96 @@
# SnailJob命名空间配置修复方案
## 问题描述
四个ruoyi-admin实例启动后都提示SnailJob Token认证失败:
```
2025-08-31 01:44:09 [snail-job-grpc-client-executor-172.16.0.67-1216] ERROR c.a.s.c.c.handler.ClientRegister
- heartbeat check requestId:[1216] message:[Token authentication failed. [namespace:prod,prod-9900 groupName:ruoyi_group_9900 token:SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT]]
```
## 问题分析
### 根本原因
1. **命名空间配置不匹配**:客户端尝试连接的命名空间是 `prod,prod-9900`,但数据库中的命名空间只有 `prod`
2. **动态配置问题**:配置文件使用 `${spring.profiles.active}` 作为命名空间,导致9900-9903实例的命名空间变成了 `prod-9900`, `prod-9901` 等
3. **数据库数据缺失**:数据库中缺少对应的命名空间记录
### 配置对比
**数据库中的命名空间:**
```sql
1 Development dev
2 Production prod
```
**客户端尝试连接的命名空间:**
- prod,prod-9900
- prod,prod-9901
- prod,prod-9902
- prod,prod-9903
## 解决方案
### 方案一:修改配置文件(已实施)
将所有实例的命名空间统一为 `prod`,而不是使用 `${spring.profiles.active}`:
**修改的文件:**
- `application-prod-9900.yml`
- `application-prod-9901.yml`
- `application-prod-9902.yml`
- `application-prod-9903.yml`
**修改内容:**
```yaml
snail-job:
namespace: "prod" # 从 ${spring.profiles.active} 改为固定值
```
### 方案二:数据库方案(备选)
如果希望保持动态命名空间配置,可以在数据库中创建对应的命名空间:
```sql
-- 为每个实例创建对应的命名空间
INSERT INTO `sj_namespace` VALUES (3, 'Production-9900', 'prod-9900', '', 0, now(), now());
INSERT INTO `sj_namespace` VALUES (4, 'Production-9901', 'prod-9901', '', 0, now(), now());
INSERT INTO `sj_namespace` VALUES (5, 'Production-9902', 'prod-9902', '', 0, now(), now());
INSERT INTO `sj_namespace` VALUES (6, 'Production-9903', 'prod-9903', '', 0, now(), now());
-- 为每个命名空间创建对应的组配置
INSERT INTO `sj_group_config` VALUES (13, 'prod-9900', 'ruoyi_group_9900', '', 'SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT', 1, 1, 0, 1, 1, now(), now());
INSERT INTO `sj_group_config` VALUES (14, 'prod-9901', 'ruoyi_group_9901', '', 'SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT', 1, 1, 0, 1, 1, now(), now());
INSERT INTO `sj_group_config` VALUES (15, 'prod-9902', 'ruoyi_group_9902', '', 'SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT', 1, 1, 0, 1, 1, now(), now());
INSERT INTO `sj_group_config` VALUES (16, 'prod-9903', 'ruoyi_group_9903', '', 'SJ_cKqBTPzCsWA3VyuCfFoccmuIEGXjr5KT', 1, 1, 0, 1, 1, now(), now());
```
## 验证步骤
1. **重启应用实例**
```bash
# 重启所有实例
./hot-start.sh
```
2. **检查日志**
- 确认不再出现Token认证失败错误
- 确认SnailJob客户端成功注册
3. **验证连接**
- 访问SnailJob管理界面
- 确认所有实例都显示为在线状态
## 最佳实践建议
1. **命名空间管理**:建议使用统一的命名空间(如 `prod`),避免动态配置导致的复杂性
2. **组名管理**:使用有意义的组名,便于识别和管理
3. **配置验证**:在部署前验证配置文件与数据库配置的一致性
4. **监控告警**:配置SnailJob连接状态的监控告警
## 相关文档
- [SnailJob官方文档](https://github.com/aizuda/snail-job)
- [项目部署文档](../部署文档/)
- [数据库设计文档](../设计文档/database/)
@@ -2652,7 +2652,7 @@ const handleAdd = async () => {
form.majorName = classData.major_name || classData.majorName;
// 确保学历层次为数字类型,避免类型不匹配导致的显示问题
const classType = classData.class_type || classData.classType;
form.academicLevel = classType ? Number(classType) : undefined;
form.academicLevel = classType ? classType : undefined;
form.grade = classData.grade;
form.enrollmentYear = classData.enrollment_year || classData.enrollmentYear;
form.classTeacherId = classData.class_teacher_id || classData.classTeacherId;