pg-test:
  hosts:
    10.10.10.11: { pg_seq: 1, pg_role: primary } # 已存在的成员
    10.10.10.12: { pg_seq: 2, pg_role: replica } # 已存在的成员
    10.10.10.13: { pg_seq: 3, pg_role: replica } # <--- 新成员
  vars: { pg_cluster: pg-test }

pg-test:
  hosts:
    10.10.10.11: { pg_seq: 1, pg_role: primary }
    10.10.10.12: { pg_seq: 2, pg_role: replica }
    10.10.10.13: { pg_seq: 3, pg_role: replica } # <--- 执行后移除此行
  vars: { pg_cluster: pg-test }

pg-test: # 清理这个集群定义分组
  hosts:
    10.10.10.11: { pg_seq: 1, pg_role: primary }
    10.10.10.12: { pg_seq: 2, pg_role: replica }
    10.10.10.13: { pg_seq: 3, pg_role: replica }
  vars: { pg_cluster: pg-test }

./pgsql-rm.yml -l pg-meta -e pg_safeguard=false    # 强制销毁受保护的 pg 集群 pg-meta
./pgsql-rm.yml -l pg-meta -e pg_rm_backup=false    # 在销毁集群过程中保留其备份仓库

./pgsql.yml -l pg-test -t pg_conf                   # 更新 Patroni 配置文件
ansible pg-test -b -a 'systemctl reload patroni'    # 重载 Patroni 服务

# pg-test 是原始集群
pg-test:
  hosts:
    10.10.10.11: { pg_seq: 1, pg_role: primary }
  vars: { pg_cluster: pg-test }

# pg-test2 是 pg-test 的备份集群（克隆）
pg-test2:
  hosts:
    10.10.10.12: { pg_seq: 1, pg_role: primary, pg_upstream: 10.10.10.11 }  # 指定上游
    10.10.10.13: { pg_seq: 2, pg_role: replica }
  vars: { pg_cluster: pg-test2 }

# 从 pg-meta 集群的备份克隆一个新集群 pg-meta2
pg-meta2:
  hosts: { 10.10.10.12: { pg_seq: 1, pg_role: primary } }
  vars:
    pg_cluster: pg-meta2
    pg_pitr:
      cluster: pg-meta                    # 从 pg-meta 的备份恢复
      time: '2025-01-10 10:00:00+00'      # 恢复到指定时间点
      archive: false                       # 独立恢复阶段禁用归档
      action: promote                      # 完成重放后提升并启动集群

pg-meta:
  hosts: { 10.10.10.10: { pg_seq: 1, pg_role: primary } }
  vars:
    pg_cluster: pg-meta
    pg_users: [{ name: dbuser_app, password: 'DBUser.App', pgbouncer: true }]  # <--- 在这里定义用户列表！

#all.children.pg-meta.vars.pg_users: # 省略上级缩进
  - name: dbuser_app
    password: DBUser.App
    pgbouncer: true
    roles: [dbrole_readwrite]
    comment: application user for myapp

- name: dbuser_app
  password: NewSecretPassword     # 修改密码

- name: dbuser_app
  superuser: false           # 超级用户（谨慎使用！）
  createdb: true             # 允许创建数据库
  createrole: false          # 允许创建角色
  inherit: true              # 自动继承角色权限
  replication: false         # 允许流复制连接
  bypassrls: false           # 绕过行级安全策略
  connlimit: 50              # 限制连接数，-1 不限制

- name: temp_user
  expire_in: 30                   # 30 天后过期（相对时间）

- name: contractor_user
  expire_at: '2024-12-31'         # 指定日期过期（绝对时间）

- name: permanent_user
  expire_at: 'infinity'           # 永不过期

- name: dbuser_app
  roles:
    - dbrole_readwrite                      # 简单形式：授予角色
    - { name: dbrole_admin, admin: true }   # 带 ADMIN OPTION（可以将此角色授予其他用户）
    - { name: pg_monitor, set: false }      # PG16+: 不允许 SET ROLE
    - { name: old_role, state: absent }     # 撤销角色成员关系

- name: dbuser_analyst
  parameters:
    work_mem: '256MB'
    statement_timeout: '5min'
    search_path: 'analytics,public'
    log_statement: DEFAULT        # 重置为默认值

- name: dbuser_app
  pgbouncer: true                 # 添加到连接池
  pool_mode: transaction          # 池化模式
  pool_connlimit: 50              # 用户最大连接数

pg_users:
  - name: dbuser_old
    state: absent

sudo -iu postgres psql -AXtwc "SELECT 1 FROM pg_roles WHERE rolname = 'dbuser_old';"
grep -n '^"dbuser_old"[[:space:]]' /etc/pgbouncer/userlist.txt

# 检查依赖关系（只读操作）
pg-drop-role dbuser_old --check

# 预览删除操作（不实际执行）
pg-drop-role dbuser_old --dry-run -v

# 确认近期备份、精确用户名与继任所有者后，才执行实际删除
pg-drop-role dbuser_old dbuser_new

# 仅当已明确同意终止连接时使用 --force
pg-drop-role dbuser_old dbuser_new --force

- name: dbuser_app
  password: DBUser.App
  pgbouncer: true
  roles: [dbrole_readwrite]
  comment: application user

- name: dbuser_readonly
  password: DBUser.Readonly
  pgbouncer: true
  roles: [dbrole_readonly]

- name: dbuser_admin
  password: DBUser.Admin
  pgbouncer: true
  pool_mode: session
  roles: [dbrole_admin]
  parameters:
    log_statement: 'all'

- name: temp_contractor
  password: TempPassword
  expire_in: 30
  roles: [dbrole_readonly]

- name: custom_role
  login: false
  comment: custom role for special permissions

- name: dbuser_special
  password: DBUser.Special
  pgbouncer: true
  roles:
    - dbrole_readwrite
    - { name: dbrole_admin, admin: true }
    - { name: pg_monitor, set: false }
    - { name: pg_execute_server_program, inherit: false }

SELECT rolname, rolsuper, rolinherit, rolcreaterole, rolcreatedb,
       rolcanlogin, rolreplication, rolbypassrls, rolconnlimit, rolvaliduntil
FROM pg_roles WHERE rolname NOT LIKE 'pg_%' ORDER BY rolname;

SELECT r.rolname AS member, g.rolname AS role, m.admin_option, m.set_option, m.inherit_option
FROM pg_auth_members m
JOIN pg_roles r ON r.oid = m.member
JOIN pg_roles g ON g.oid = m.roleid
WHERE r.rolname = 'dbuser_app';

SELECT rolname, setconfig FROM pg_db_role_setting s
JOIN pg_roles r ON r.oid = s.setrole WHERE s.setdatabase = 0;

SELECT rolname, rolvaliduntil, rolvaliduntil - CURRENT_TIMESTAMP AS time_remaining
FROM pg_roles WHERE rolvaliduntil IS NOT NULL
  AND rolvaliduntil < CURRENT_TIMESTAMP + INTERVAL '30 days'
ORDER BY rolvaliduntil;

# 在目标集群当前主库上，交互修改数据库角色密码
sudo -iu postgres psql -d postgres
\password dbuser_dba       # 或 dbuser_monitor / replicator

# 管理员 dbuser_dba：PG 节点 .pgpass、Pgbouncer、Infra 管理端与 pgAdmin 文件
./pgsql.yml -l <cls> -t pg_pass,pgbouncer_user,pgbouncer_reload -e pg_reload=true
./infra.yml -l infra -t env_pgpass,env_pgscv,env_pgadmin

# 监控用户 dbuser_monitor：PG 节点 .pgpass、Pgbouncer、两个 Exporter 与 Grafana 数据源
./pgsql.yml -l <cls> -t pg_pass,pgbouncer_user,pgbouncer_reload,pg_exporter,pgbouncer_exporter,add_ds -e pg_reload=true
./infra.yml -l infra -t env_pgpass

# 复制用户 replicator：Patroni 配置、PG 节点 .pgpass 与 Infra .pgpass
./pgsql.yml -l <cls> -t pg_conf,pg_pass,patroni_reload -e pg_reload=true
./infra.yml -l infra -t env_pgpass

./pgsql.yml -l <cls> -t pg_conf,patroni_reload -e pg_reload=true
./infra.yml -l infra -t env_patroni

pg-meta:
  hosts: { 10.10.10.10: { pg_seq: 1, pg_role: primary } }
  vars:
    pg_cluster: pg-meta
    pg_databases: [{ name: some_db }]  # <--- 在这里定义数据库列表！

#all.children.pg-meta.vars.pg_databases: # 省略上级缩进
  - name: myapp
    owner: dbuser_myapp
    schemas: [app]
    extensions:
      - { name: pg_trgm }
      - { name: btree_gin }
    comment: my application database

- name: myapp
  owner: dbuser_new_owner     # 修改为新属主

- name: myapp
  connlimit: 100              # 限制最大 100 个连接

- name: myapp
  owner: dbuser_myapp
  revokeconn: true            # 回收 PUBLIC 的 CONNECT 权限

- name: myapp
  parameters:
    work_mem: '256MB'
    maintenance_work_mem: '512MB'
    statement_timeout: '30s'
    search_path: DEFAULT      # 重置为默认值

- name: myapp
  schemas:
    - app                                   # 简单形式
    - { name: core, owner: dbuser_myapp }   # 指定属主
    - { name: deprecated, state: absent }   # 删除模式

- name: myapp
  extensions:
    - postgis                                 # 简单形式
    - { name: vector, schema: public }        # 指定 schema
    - { name: pg_trgm, state: absent }        # 卸载扩展

- name: myapp
  pgbouncer: true              # 是否加入连接池（默认 true）
  pool_mode: transaction       # 池化模式：transaction/session/statement
  pool_size: 50                # 默认池大小
  pool_reserve: 30             # 保留池大小
  pool_size_min: 0             # 最小池大小
  pool_connlimit: 100          # 最大数据库连接
  pool_auth_user: dbuser_meta  # 认证查询使用用户（配合 pgbouncer_auth_query）

pg_databases:
  - name: olddb
    state: absent

pg_databases:
  - name: testdb
    state: recreate
    owner: dbuser_test
    baseline: test_init.sql    # 重建后执行初始化

pg_databases:
  - name: meta                   # 源数据库

  - name: meta_dev
    template: meta               # 以 meta 作为模板
    strategy: FILE_COPY          # PG15+ 克隆策略，PG18 瞬间生效

SELECT pg_terminate_backend(pid) FROM pg_stat_activity WHERE datname = 'meta';
CREATE DATABASE meta_dev TEMPLATE meta STRATEGY FILE_COPY;

pg edit-config <cls>                  # 交互式编辑集群配置
pg edit-config <cls> --force          # 跳过确认提示直接应用
pg edit-config <cls> -p <k>=<v>       # 修改 PostgreSQL 参数（--pg 简写）
pg edit-config <cls> -s <k>=<v>       # 修改 Patroni 参数（--set 简写）

# 修改 PostgreSQL 参数：慢查询阈值（会询问是否应用）
pg edit-config pg-test -p log_min_duration_statement=1000

# 修改 PostgreSQL 参数并跳过确认
pg edit-config pg-test -p log_min_duration_statement=1000 --force

# 修改多个 PostgreSQL 参数
pg edit-config pg-test -p work_mem=256MB -p maintenance_work_mem=1GB --force

# 修改 Patroni 参数：增大故障检测时间窗口（增大 RTO）
pg edit-config pg-test -s loop_wait=15 -s ttl=60 --force

# 修改 Patroni 参数：启用同步复制模式
pg edit-config pg-test -s synchronous_mode=true --force

# 修改 Patroni 参数：启用严格同步模式（至少一个同步从库才允许写入）
pg edit-config pg-test -s synchronous_mode_strict=true --force

# 修改需要重启的参数（修改后需执行 pg restart）
pg edit-config pg-test -p shared_buffers=4GB --force
pg edit-config pg-test -p shared_preload_libraries='timescaledb, pg_stat_statements' --force
pg edit-config pg-test -p max_connections=200 --force

# 查看当前配置
curl -s 10.10.10.11:8008/config | jq .

# 通过 API 修改参数（需要认证）
curl -u 'postgres:Patroni.API' \
     -d '{"postgresql":{"parameters": {"log_min_duration_statement":200}}}' \
     -s -X PATCH http://10.10.10.11:8008/config | jq .

pg list <cls>                         # 查看指定集群的状态
pg list                               # 列出所有集群（需要在管理节点上执行）
pg list <cls> -e                      # 显示扩展信息（--extended）
pg list <cls> -t                      # 显示时间戳（--timestamp）
pg list <cls> -f json                 # 以 JSON 格式输出（--format）
pg list <cls> -W 5                    # 每 5 秒刷新一次（--watch）

+ Cluster: pg-test (7322261897169354773) -----+----+--------------+
| Member    | Host        | Role    | State   | TL | Lag in MB    |
+-----------+-------------+---------+---------+----+--------------+
| pg-test-1 | 10.10.10.11 | Leader  | running |  1 |              |
| pg-test-2 | 10.10.10.12 | Replica | running |  1 |            0 |
| pg-test-3 | 10.10.10.13 | Replica | running |  1 |            0 |
+-----------+-------------+---------+---------+----+--------------+

+ Cluster: pg-test (7322261897169354773) -------+----+--------------+
| Member      | Host        | Role    | State   | TL | Lag in MB    |
+-------------+-------------+---------+---------+----+--------------+
| pg-test-1 * | 10.10.10.11 | Leader  | running |  1 |              |
| pg-test-2 * | 10.10.10.12 | Replica | running |  1 |            0 |
+-------------+-------------+---------+---------+----+--------------+

pg switchover <cls>                   # 交互式切换，会提示选择目标从库
pg switchover <cls> --leader <old>    # 指定当前主库名称
pg switchover <cls> --candidate <new> # 指定目标从库名称
pg switchover <cls> --scheduled <time> # 定时切换，格式如 2024-12-01T03:00
pg switchover <cls> --force           # 跳过确认提示

# 交互式切换（推荐，会显示当前拓扑并提示选择）
$ pg switchover pg-test
Current cluster topology
+ Cluster: pg-test (7322261897169354773) -----+----+--------------+
| Member    | Host        | Role    | State   | TL | Lag in MB    |
+-----------+-------------+---------+---------+----+--------------+
| pg-test-1 | 10.10.10.11 | Leader  | running |  1 |              |
| pg-test-2 | 10.10.10.12 | Replica | running |  1 |            0 |
| pg-test-3 | 10.10.10.13 | Replica | running |  1 |            0 |
+-----------+-------------+---------+---------+----+--------------+
Primary [pg-test-1]:
Candidate ['pg-test-2', 'pg-test-3'] []: pg-test-2
When should the switchover take place (e.g. 2024-01-01T12:00) [now]:
Are you sure you want to switchover cluster pg-test, demoting current leader pg-test-1? [y/N]: y

# 非交互式切换（指定主库和候选从库）
pg switchover pg-test --leader pg-test-1 --candidate pg-test-2 --force

# 定时切换（在凌晨 3 点执行，适合维护窗口）
pg switchover pg-test --leader pg-test-1 --candidate pg-test-2 --scheduled "2024-12-01T03:00"

pg failover <cls>                     # 交互式故障切换
pg failover <cls> --candidate <new>   # 指定要提升的从库
pg failover <cls> --force             # 跳过确认提示

# 交互式故障切换
$ pg failover pg-test
Candidate ['pg-test-2', 'pg-test-3'] []: pg-test-2
Are you sure you want to failover cluster pg-test? [y/N]: y
Successfully failed over to "pg-test-2"

# 非交互式故障切换（紧急情况快速执行）
pg failover pg-test --candidate pg-test-2 --force

pg restart <cls>                      # 重启整个集群的所有实例
pg restart <cls> <member>             # 重启指定实例
pg restart <cls> --role leader        # 仅重启主库
pg restart <cls> --role replica       # 仅重启所有从库
pg restart <cls> --pending            # 仅重启标记为需要重启的实例
pg restart <cls> --scheduled <time>   # 定时重启
pg restart <cls> --timeout <sec>      # 设置重启超时时间（秒）
pg restart <cls> --force              # 跳过确认提示

# 查看哪些实例需要重启（名称后带 * 标记）
$ pg list pg-test
+ Cluster: pg-test (7322261897169354773) -------+----+--------------+
| Member      | Host        | Role    | State   | TL | Lag in MB    |
+-------------+-------------+---------+---------+----+--------------+
| pg-test-1 * | 10.10.10.11 | Leader  | running |  1 |              |
| pg-test-2 * | 10.10.10.12 | Replica | running |  1 |            0 |
+-------------+-------------+---------+---------+----+--------------+

# 重启单个从库实例
pg restart pg-test pg-test-2

# 重启整个集群的所有成员（不承诺 leader-last 顺序）
pg restart pg-test --force

# 仅重启需要重启的实例
pg restart pg-test --pending --force

# 显式按“从库优先、主库最后”执行
pg restart pg-test --role replica --force
pg restart pg-test --role leader --force

# 定时重启（在维护窗口执行）
pg restart pg-test --scheduled "2024-12-01T03:00"

# 设置重启超时时间为 300 秒
pg restart pg-test --timeout 300 --force

pg reload <cls>                       # 重载整个集群的配置
pg reload <cls> <member>              # 重载指定实例的配置
pg reload <cls> --role leader         # 仅重载主库
pg reload <cls> --role replica        # 仅重载所有从库
pg reload <cls> --force               # 跳过确认提示

# 重载整个集群
pg reload pg-test

# 重载单个实例
pg reload pg-test pg-test-1

# 强制重载，跳过确认
pg reload pg-test --force

pg reinit <cls> <member>              # 重新初始化指定从库
pg reinit <cls> <member> --force      # 跳过确认提示
pg reinit <cls> <member> --wait       # 等待重建完成后再返回

# 重新初始化从库（会提示确认）
$ pg reinit pg-test pg-test-2
Are you sure you want to reinitialize members pg-test-2? [y/N]: y
Success: reinitialize for member pg-test-2

# 强制重新初始化，跳过确认
pg reinit pg-test pg-test-2 --force

# 重新初始化并等待完成
pg reinit pg-test pg-test-2 --force --wait

+ Cluster: pg-test (7322261897169354773) --------------+----+------+
| Member    | Host        | Role    | State            | TL | Lag  |
+-----------+-------------+---------+------------------+----+------+
| pg-test-1 | 10.10.10.11 | Leader  | running          |  2 |      |
| pg-test-2 | 10.10.10.12 | Replica | creating replica |    |    ? |
+-----------+-------------+---------+------------------+----+------+

pg pause <cls>                        # 暂停自动故障转移
pg pause <cls> --wait                 # 暂停并等待所有成员确认

# 暂停自动切换
$ pg pause pg-test
Success: cluster management is paused

# 查看集群状态（底部会显示 Maintenance mode: on）
$ pg list pg-test
+ Cluster: pg-test (7322261897169354773) -----+----+--------------+
| Member    | Host        | Role    | State   | TL | Lag in MB    |
+-----------+-------------+---------+---------+----+--------------+
| pg-test-1 | 10.10.10.11 | Leader  | running |  1 |              |
| pg-test-2 | 10.10.10.12 | Replica | running |  1 |            0 |
+-----------+-------------+---------+---------+----+--------------+
 Maintenance mode: on

pg resume <cls>                       # 恢复自动故障转移
pg resume <cls> --wait                # 恢复并等待所有成员确认

# 恢复自动切换
$ pg resume pg-test
Success: cluster management is resumed

# 确认已恢复（Maintenance mode 提示消失）
$ pg list pg-test

pg history <cls>                      # 显示故障转移历史
pg history <cls> -f json              # 以 JSON 格式输出
pg history <cls> -f yaml              # 以 YAML 格式输出

$ pg history pg-test
+----+-----------+------------------------------+---------------------------+
| TL |       LSN | Reason                       | Timestamp                 |
+----+-----------+------------------------------+---------------------------+
|  1 | 0/5000060 | no recovery target specified | 2024-01-15T10:30:00+08:00 |
|  2 | 0/6000000 | switchover to pg-test-2      | 2024-01-20T14:00:00+08:00 |
|  3 | 0/7000028 | failover to pg-test-1        | 2024-01-25T09:15:00+08:00 |
+----+-----------+------------------------------+---------------------------+

pg show-config <cls>                  # 显示集群配置

$ pg show-config pg-test
loop_wait: 10
maximum_lag_on_failover: 1048576
postgresql:
  parameters:
    archive_command: pgbackrest --stanza=pg-test archive-push %p
    max_connections: 100
    shared_buffers: 256MB
    log_min_duration_statement: 1000
  use_pg_rewind: true
  use_slots: true
retry_timeout: 10
ttl: 30
synchronous_mode: false

pg query <cls> -c "<sql>"             # 在主库上执行查询
pg query <cls> -c "<sql>" -m <member> # 在指定实例上执行（--member）
pg query <cls> -c "<sql>" -r leader   # 在主库上执行（--role）
pg query <cls> -c "<sql>" -r replica  # 在所有从库上执行
pg query <cls> -f <file>              # 从文件读取 SQL 执行
pg query <cls> -c "<sql>" -U <user>   # 指定用户名（--username）
pg query <cls> -c "<sql>" -d <db>     # 指定数据库（--dbname）
pg query <cls> -c "<sql>" --format json  # 以 JSON 格式输出

# 查看主库当前连接数
pg query pg-test -c "SELECT count(*) FROM pg_stat_activity"

# 查看 PostgreSQL 版本
pg query pg-test -c "SELECT version()"

# 在所有从库上查看复制状态
pg query pg-test -c "SELECT pg_is_in_recovery(), pg_last_wal_replay_lsn()" -r replica

# 在指定实例上执行
pg query pg-test -c "SELECT pg_is_in_recovery()" -m pg-test-2

# 使用指定用户和数据库
pg query pg-test -c "SELECT current_user, current_database()" -U postgres -d postgres

# 以 JSON 格式输出结果
pg query pg-test -c "SELECT * FROM pg_stat_replication" --format json

pg topology <cls>                     # 显示复制拓扑

$ pg topology pg-test
+ Cluster: pg-test (7322261897169354773) -------+----+--------------+
| Member      | Host        | Role    | State   | TL | Lag in MB    |
+-------------+-------------+---------+---------+----+--------------+
| pg-test-1   | 10.10.10.11 | Leader  | running |  1 |              |
| + pg-test-2 | 10.10.10.12 | Replica | running |  1 |            0 |
| + pg-test-3 | 10.10.10.13 | Replica | running |  1 |            0 |
+-------------+-------------+---------+---------+----+--------------+

pg version                            # 显示 patronictl 版本

$ pg version
patronictl version 4.1.0

pg remove <cls>                       # 从 DCS 中移除整个集群的元数据

# 移除整个集群的元数据（需要多次确认）
$ pg remove pg-test
Please confirm the cluster name to remove: pg-test
You are about to remove all information in DCS for pg-test, please type: "Yes I am aware": Yes I am aware

pg-meta:
  hosts: { 10.10.10.10: { pg_seq: 1, pg_role: primary } }
  vars:
    pg_cluster: pg-meta
    pg_hba_rules:                            # <--- 在这里定义 HBA 规则列表！
      - {user: dbuser_app, db: app, addr: intra, auth: pwd, title: 'app access'}
      - {user: dbuser_api, db: all, addr: world, auth: ssl, title: 'api ssl access'}

# 重载配置（会验证语法）
psql -c "SELECT pg_reload_conf()"

# 如果有语法错误，查看日志
tail -f /pg/log/postgresql-*.log

pg-meta:
  vars:
    pg_hba_rules:
      - {user: new_user, db: new_db, addr: '192.168.1.0/24', auth: pwd, title: 'new app access'}

bin/pgsql-hba pg-meta

pg_hba_rules:
  - {user: all, db: all, addr: '10.1.1.100/32', auth: deny, order: 0, title: 'emergency block'}

bin/pgsql-hba pg-meta    # 立即刷新生效

pg_hba_rules:
  # 仅主库允许写入用户
  - {user: writer, db: all, addr: intra, auth: pwd, role: primary, title: 'writer on primary'}
  # 从库允许只读用户
  - {user: reader, db: all, addr: world, auth: ssl, role: replica, title: 'reader on replica'}

./pgsql.yml -l 10.10.10.14       # 扩容新实例
bin/pgsql-hba pg-meta            # 刷新所有实例的 HBA（包含新成员 IP）

# 更新 pigsty.yml 中的角色定义后刷新
bin/pgsql-hba pg-meta

psql -c "TABLE pg_hba_file_rules"

bin/pgsql-hba <cls>

SELECT * FROM pg_roles WHERE rolname = 'xxx';

psql -c "SELECT pg_reload_conf()"

head -20 /pg/data/pg_hba.conf

pg_databases:
  - name: mydb                # 默认加入连接池
    pool_auth_user: dbuser_meta # 可选，认证查询用户（配合 pgbouncer_auth_query）
    pool_mode: transaction    # 数据库级池化模式
    pool_size: 50             # 默认池大小
    pool_reserve: 30          # 保留池大小
    pool_size_min: 0          # 最小池大小
    pool_connlimit: 100       # 最大数据库连接数
  - name: internal
    pgbouncer: false          # 不加入连接池

pg_users:
  - name: dbuser_app
    password: DBUser.App
    pgbouncer: true           # 加入连接池用户列表
    pool_mode: transaction    # 用户级池化模式
    pool_connlimit: 50        # 用户级最大连接数

# 编辑配置
$ vim /etc/pgbouncer/pgbouncer.ini

# 重载生效：通过 systemctl
$ sudo systemctl reload pgbouncer

# 重载生效，本身是 pg_dbsu / postgres 用户
$ pgb -c "RELOAD;"

alias pgb='psql -p6432 -dpgbouncer'

$ pgb
pgbouncer=# SHOW POOLS;
pgbouncer=# SHOW CLIENTS;
pgbouncer=# SHOW SERVERS;

PAUSE [db];           -- 暂停指定数据库，不指定则暂停所有数据库

$ pgb -c "PAUSE mydb;"        # 暂停 mydb 数据库
$ pgb -c "PAUSE;"             # 暂停所有数据库

pgbouncer=# SHOW DATABASES;
   name   |   host    | port | database | ... | paused | disabled
----------+-----------+------+----------+-----+--------+----------
 mydb     | /var/run  | 5432 | mydb     | ... |      1 |        0

RESUME [db];          -- 恢复指定数据库，不指定则恢复所有数据库

$ pgb -c "RESUME mydb;"       # 恢复 mydb 数据库
$ pgb -c "RESUME;"            # 恢复所有数据库

DISABLE db;           -- 禁用指定数据库（必须指定数据库名）

$ pgb -c "DISABLE mydb;"      # 禁用 mydb，新连接被拒绝

ENABLE db;            -- 启用指定数据库（必须指定数据库名）

$ pgb -c "ENABLE mydb;"       # 启用 mydb，允许新连接

RECONNECT [db];       -- 重建指定数据库的服务端连接，不指定则重建所有

$ pgb -c "RECONNECT mydb;"    # 重建 mydb 的服务端连接
$ pgb -c "RECONNECT;"         # 重建所有服务端连接

KILL [db];            -- 终止指定数据库的所有连接，不指定则终止所有（admin 除外）

$ pgb -c "KILL mydb;"         # 强制断开 mydb 的所有连接
$ pgb -c "KILL;"              # 强制断开所有数据库的连接（admin 除外）

KILL_CLIENT id;       -- 终止指定 ID 的客户端连接

# 查看客户端连接
$ pgb -c "SHOW CLIENTS;"

# 终止特定客户端（假设 ptr 列显示的 ID 为 0x1234567890）
$ pgb -c "KILL_CLIENT 0x1234567890;"

SUSPEND;              -- 挂起 Pgbouncer

# 1. 挂起当前 Pgbouncer
$ pgb -c "SUSPEND;"

# 2. 启动新的 Pgbouncer 进程（使用 -R 选项接管 socket）
$ pgbouncer -R /etc/pgbouncer/pgbouncer.ini

# 3. 新进程接管后，旧进程自动退出

SHUTDOWN;                      -- 立即关闭
SHUTDOWN WAIT_FOR_SERVERS;     -- 等待服务端连接释放后关闭
SHUTDOWN WAIT_FOR_CLIENTS;     -- 等待客户端断开后关闭（零停机滚动重启）

$ pgb -c "SHUTDOWN WAIT_FOR_CLIENTS;"   # 优雅关闭，等待客户端断开

RELOAD;               -- 重载配置文件

$ pgb -c "RELOAD;"              # 通过管理控制台重载
$ systemctl reload pgbouncer    # 通过 systemd 重载
$ kill -SIGHUP $(cat /run/postgresql/pgbouncer.pid)  # 通过信号重载

./pgsql.yml -l <cls> -t pgbouncer_reload    # 重载集群的 Pgbouncer 配置

WAIT_CLOSE [db];      -- 等待指定数据库的服务端连接关闭，不指定则等待所有

# 完整的连接重建流程
$ pgb -c "RECONNECT mydb;"
$ pgb -c "WAIT_CLOSE mydb;"    # 等待旧连接释放

# 查看连接池状态
$ pgb -c "SHOW POOLS;"

# 查看客户端连接
$ pgb -c "SHOW CLIENTS;"

# 查看服务端连接
$ pgb -c "SHOW SERVERS;"

# 查看统计信息
$ pgb -c "SHOW STATS;"

# 查看数据库状态
$ pgb -c "SHOW DATABASES;"

# 通过信号重载配置
$ kill -SIGHUP $(cat /run/postgresql/pgbouncer.pid)

# 通过信号优雅关闭
$ kill -SIGTERM $(cat /run/postgresql/pgbouncer.pid)

# 通过信号暂停
$ kill -SIGUSR1 $(cat /run/postgresql/pgbouncer.pid)

# 通过信号恢复
$ kill -SIGUSR2 $(cat /run/postgresql/pgbouncer.pid)

# 1. 仅把 mydb 的后端目标改为 10.10.10.12
$ sed -i -E '/^mydb[[:space:]]*=/ s#host=[^[:space:]]+#host=10.10.10.12#' /etc/pgbouncer/database.txt

# 2. 重载配置
$ pgb -c "RELOAD;"

# 3. 重建该数据库的连接并等待旧连接释放
$ pgb -c "RECONNECT mydb;"
$ pgb -c "WAIT_CLOSE mydb;"

systemctl start patroni     # 启动 Patroni（同时启动 PostgreSQL）

systemctl stop patroni      # 停止 Patroni（同时停止 PostgreSQL）

systemctl restart patroni   # 重启 Patroni（同时重启 PostgreSQL）

systemctl reload patroni    # 重载 Patroni 配置

systemctl status patroni    # 查看 Patroni 服务状态
journalctl -u patroni -f    # 实时查看 Patroni 日志
journalctl -u patroni -n 100 --no-pager  # 查看最近 100 行日志

systemctl start pgbouncer

systemctl stop pgbouncer

systemctl restart pgbouncer

systemctl reload pgbouncer

systemctl status pgbouncer
journalctl -u pgbouncer -f

psql -p 6432 -U postgres -d pgbouncer  # 连接到 Pgbouncer 管理控制台

SHOW POOLS;      -- 查看连接池状态
SHOW CLIENTS;    -- 查看客户端连接
SHOW SERVERS;    -- 查看后端服务器连接
SHOW STATS;      -- 查看统计信息
RELOAD;          -- 重载配置
PAUSE;           -- 暂停所有连接池
RESUME;          -- 恢复所有连接池

systemctl start haproxy

systemctl stop haproxy

systemctl restart haproxy

systemctl reload haproxy

systemctl status haproxy
journalctl -u haproxy -f

http://<node_ip>:9101/haproxy

systemctl start pg_exporter

systemctl stop pg_exporter

systemctl restart pg_exporter

systemctl status pg_exporter
journalctl -u pg_exporter -f

curl -s localhost:9630/metrics | head -20

systemctl start pgbouncer_exporter
systemctl stop pgbouncer_exporter
systemctl restart pgbouncer_exporter

systemctl status pgbouncer_exporter
journalctl -u pgbouncer_exporter -f

curl -s localhost:9631/metrics | head -20

systemctl start vip-manager

systemctl stop vip-manager

systemctl restart vip-manager

systemctl status vip-manager
journalctl -u vip-manager -f

ip addr show           # 查看网络接口，检查 VIP 是否绑定
pg list <cls>          # 确认主库位置

1. patroni          # 首先启动 Patroni（会自动启动 PostgreSQL）
2. pgbouncer        # 然后启动连接池
3. haproxy          # 启动负载均衡器
4. pg_exporter      # 启动监控导出器
5. pgbouncer_exporter
6. vip-manager      # 最后启动 VIP 管理器（如果启用）

systemctl start patroni pgbouncer haproxy pg_exporter pgbouncer_exporter

systemctl stop pgbouncer_exporter pg_exporter haproxy pgbouncer patroni

systemctl status <service>        # 查看服务状态
journalctl -u <service> -n 50     # 查看最近日志
journalctl -u <service> --since "5 min ago"  # 查看最近 5 分钟日志