典籍篇
聚社区之智,传技术之火
源码安装PostgreSQL稳定版本(Linux)
📋 准备工作
1. 查看最新稳定版本
# 访问PostgreSQL官网或使用命令查看
curl -s https://www.postgresql.org/ftp/source/ | grep -E 'href="v[0-9]+\.[0-9]+/"' | tail -5
# 当前长期支持版本(LTS)
# PostgreSQL 16.x (最新稳定版)
# PostgreSQL 15.x (长期支持)
# PostgreSQL 14.x (长期支持到2026年)
# PostgreSQL 13.x (长期支持到2025年)
2. 安装编译依赖
# Ubuntu/Debian
sudo apt update
sudo apt install -y \
build-essential \
libreadline-dev \
zlib1g-dev \
flex bison \
libxml2-dev \
libxslt-dev \
libssl-dev \
libssl-dev \
libpam0g-dev \
libldap-dev \
libperl-dev \
libicu-dev \
tcl-dev \
python3-dev \
git \
wget \
curl
# CentOS/RHEL/Rocky/AlmaLinux
sudo yum groupinstall -y "Development Tools"
sudo yum install -y \
readline-devel \
zlib-devel \
flex bison \
libxml2-devel \
libxslt-devel \
openssl-devel \
pam-devel \
openldap-devel \
perl-devel \
perl-ExtUtils-Embed \
tcl-devel \
python3-devel \
git \
wget
vi /etc/sysctl.conf
#最大共享内存段大小
kernel.shmmax = 68719476736(默认)
#可以使用的共享内存的总量
kernel.shmall = 4294967296(默认)
#整个系统共享内存段的最大数目
kernel.shmmni = 4096
#每个信号对象集的最大信号对象数
kernel.sem = 50100 64128000 50100 1280
#文件句柄的最大数量。
fs.file-max = 7672460
#应用程序可使用的IPv4端口范围
net.ipv4.ip_local_port_range = 9000 65000
#套接字接收缓冲区大小的缺省值
net.core.rmem_default = 1048576
#套接字发送缓冲区大小的缺省
net.core.wmem_default = 262144 值
#套接字发送缓冲区大小的最大值
net.core.wmem_max = 1048576
# sysctl -p #配置生效
🚀 源码安装PostgreSQL 16.2(当前稳定版)
步骤1:创建专用用户和目录
# 创建postgres系统用户和组
sudo groupadd -r postgres
sudo useradd -r -g postgres -s /bin/bash -d /usr/local/pgsql -m -k /dev/null postgres
# 创建安装目录和数据目录
sudo mkdir -p /usr/local/pgsql/{data,logs,backup}
sudo chown -R postgres:postgres /usr/local/pgsql
sudo chmod 750 /usr/local/pgsql
# 创建源码目录
sudo mkdir -p /opt/postgresql_src
sudo chown $(whoami):$(whoami) /opt/postgresql_src
cd /opt/postgresql_src
步骤2:下载源码
# 下载PostgreSQL 16.2(截至2024年3月的最新稳定版)
wget https://ftp.postgresql.org/pub/source/v16.2/postgresql-16.2.tar.gz
# 或者使用国内镜像(清华源)
# wget https://mirrors.tuna.tsinghua.edu.cn/postgresql/source/v16.2/postgresql-16.2.tar.gz
# 验证下载完整性(可选)
wget https://ftp.postgresql.org/pub/source/v16.2/postgresql-16.2.tar.gz.sha256
sha256sum -c postgresql-16.2.tar.gz.sha256
# 解压源码
tar -zxvf postgresql-16.2.tar.gz
cd postgresql-16.2
步骤3:配置编译选项
# 查看所有配置选项
./configure --help
# 创建编译目录(推荐)
mkdir build && cd build
# 基本配置(适用于大多数情况)
../configure \
--prefix=/usr/local/pgsql \
--with-pgport=5432 \
--with-perl \
--with-python \
--with-tcl \
--with-openssl \
--with-pam \
-with-uuid=ossp \
--with-ldap \
--with-libxml \
--with-libxslt \
--with-icu \
--enable-thread-safety \
--enable-debug \
--enable-nls \
--with-system-tzdata=/usr/share/zoneinfo
# 或者精简配置(最小化安装)
# ../configure --prefix=/usr/local/pgsql --with-openssl
步骤4:编译和安装
# 查看CPU核心数,决定并行编译数
nproc
# 编译(使用4个并行任务,根据CPU核心数调整)
make -j4
# 可选:运行回归测试(需要较长时间)
# make check
# 安装到系统
sudo make install
# 安装contrib模块(扩展工具)
cd contrib
make -j4
sudo make install
cd ..
步骤5:配置环境变量
# 为postgres用户配置环境变量
sudo tee -a /usr/local/pgsql/.bash_profile << 'EOF'
export PGHOME=/usr/local/pgsql
export PGDATA=/usr/local/pgsql/data
export PATH=$PGHOME/bin:$PATH
export LD_LIBRARY_PATH=$PGHOME/lib:$LD_LIBRARY_PATH
export MANPATH=$PGHOME/share/man:$MANPATH
EOF
# 为当前用户配置环境变量
echo 'export PATH=/usr/local/pgsql/bin:$PATH' >> ~/.bashrc
source ~/.bashrc
# 验证安装
/usr/local/pgsql/bin/postgres --version
⚙️ 初始化数据库和配置
步骤6:初始化数据库
# 切换到postgres用户
sudo su - postgres
# 初始化数据库集群
initdb -D /usr/local/pgsql/data -E UTF8 --locale=C -U postgres
# 或者使用详细参数
initdb \
-D /usr/local/pgsql/data \
-E UTF8 \
--locale=en_US.UTF-8 \
--lc-collate=C \
--lc-ctype=en_US.UTF-8 \
--username=postgres \
--pwprompt
# 设置密码(记下这个密码)
# 输入并确认postgres用户的密码
步骤7:配置postgresql.conf
# 备份原始配置
cp /usr/local/pgsql/data/postgresql.conf /usr/local/pgsql/data/postgresql.conf.original
# 编辑主配置文件
nano /usr/local/pgsql/data/postgresql.conf
# 修改以下参数(根据服务器配置调整)
listen_addresses = '*' # 允许远程连接
port = 5432 # 监听端口
max_connections = 100 # 最大连接数
shared_buffers = 128MB # 共享缓冲区大小(建议为内存的25%)
work_mem = 4MB # 每个查询的工作内存
maintenance_work_mem = 64MB # 维护操作的内存
dynamic_shared_memory_type = posix # 动态共享内存类型
wal_level = replica # WAL级别
synchronous_commit = on # 同步提交
wal_buffers = -1 # WAL缓冲区(-1表示自动)
checkpoint_timeout = 5min # 检查点超时时间
max_wal_size = 1GB # 最大WAL大小
min_wal_size = 80MB # 最小WAL大小
archive_mode = off # 归档模式
archive_command = '/bin/date' # 归档命令(关闭时随意设置)
logging_collector = on # 启用日志收集
log_directory = 'pg_log' # 日志目录
log_filename = 'postgresql-%Y-%m-%d_%H%M%S.log' # 日志文件名格式
log_rotation_age = 1d # 日志轮转时间
log_rotation_size = 10MB # 日志轮转大小
log_truncate_on_rotation = on # 轮转时截断
log_line_prefix = '%m [%p] %q%u@%d ' # 日志前缀
log_timezone = 'Asia/Shanghai' # 日志时区
timezone = 'Asia/Shanghai' # 时区
datestyle = 'iso, ymd' # 日期格式
lc_messages = 'en_US.UTF-8' # 消息语言
lc_monetary = 'en_US.UTF-8' # 货币格式
lc_numeric = 'en_US.UTF-8' # 数字格式
lc_time = 'en_US.UTF-8' # 时间格式
default_text_search_config = 'pg_catalog.english' # 全文搜索配置
步骤8:配置客户端认证
# 编辑pg_hba.conf
nano /usr/local/pgsql/data/pg_hba.conf
# 允许本地连接(md5需要密码,trust不需要密码)
local all postgres peer
local all all md5
host all all 127.0.0.1/32 md5
# 允许特定IP段连接
host all all 192.168.1.0/24 md5
# 允许所有IP连接(生产环境慎用)
host all all 0.0.0.0/0 md5
# 示例:只允许特定用户从特定IP访问特定数据库
# host exam_db app_user 192.168.1.100/32 md5
步骤9:创建系统服务
# 退出postgres用户
exit
# 创建systemd服务文件
sudo tee /etc/systemd/system/postgresql.service << 'EOF'
[Unit]
Description=PostgreSQL Database Server
After=network.target
[Service]
Type=forking
User=postgres
Group=postgres
Environment=PGDATA=/usr/local/pgsql/data
OOMScoreAdjust=-1000
ExecStart=/usr/local/pgsql/bin/pg_ctl -D ${PGDATA} start
ExecStop=/usr/local/pgsql/bin/pg_ctl -D ${PGDATA} stop
ExecReload=/usr/local/pgsql/bin/pg_ctl -D ${PGDATA} reload
TimeoutSec=300
[Install]
WantedBy=multi-user.target
EOF
# 或者使用simple类型(推荐)
sudo tee /etc/systemd/system/postgresql.service << 'EOF'
[Unit]
Description=PostgreSQL Database Server
After=network.target
[Service]
Type=simple
User=postgres
Group=postgres
Environment=PGDATA=/usr/local/pgsql/data
Environment=PGPORT=5432
OOMScoreAdjust=-1000
ExecStart=/usr/local/pgsql/bin/postgres -D ${PGDATA}
ExecReload=/bin/kill -HUP $MAINPID
KillMode=mixed
KillSignal=SIGINT
TimeoutSec=300
[Install]
WantedBy=multi-user.target
EOF
步骤10:启动服务
# 重新加载systemd配置
sudo systemctl daemon-reload
# 启动PostgreSQL服务
sudo systemctl start postgresql
# 设置开机自启
sudo systemctl enable postgresql
# 查看服务状态
sudo systemctl status postgresql
# 查看日志
sudo journalctl -u postgresql -f
🔧 验证和基础配置
步骤11:验证安装
# 连接到数据库
psql -h localhost -U postgres -d postgres
# 在psql中执行
SELECT version();
SELECT current_user;
SHOW data_directory;
SHOW config_file;
\q # 退出
步骤12:创建数据库和用户
# 切换到postgres用户
sudo su - postgres
# 创建测试数据库
createdb test_db
# 创建应用用户
createuser -P -d -e app_user
# 输入密码:AppPass123
# 连接到数据库并授权
psql -d test_db <<EOF
-- 创建schema
CREATE SCHEMA IF NOT EXISTS app_schema;
-- 创建测试表
CREATE TABLE app_schema.users (
id SERIAL PRIMARY KEY,
username VARCHAR(50) NOT NULL,
email VARCHAR(100) UNIQUE NOT NULL,
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);
-- 插入测试数据
INSERT INTO app_schema.users (username, email) VALUES
('admin', 'admin@example.com'),
('user1', 'user1@example.com');
-- 授权给应用用户
GRANT ALL PRIVILEGES ON DATABASE test_db TO app_user;
GRANT ALL PRIVILEGES ON SCHEMA app_schema TO app_user;
GRANT ALL PRIVILEGES ON ALL TABLES IN SCHEMA app_schema TO app_user;
GRANT USAGE, SELECT ON ALL SEQUENCES IN SCHEMA app_schema TO app_user;
-- 查看表
SELECT * FROM app_schema.users;
EOF
步骤13:安装常用扩展
# 安装pg_stat_statements(性能监控)
psql -U postgres -d test_db <<EOF
CREATE EXTENSION IF NOT EXISTS pg_stat_statements;
CREATE EXTENSION IF NOT EXISTS uuid-ossp;
CREATE EXTENSION IF NOT EXISTS btree_gin;
CREATE EXTENSION IF NOT EXISTS btree_gist;
SELECT * FROM pg_available_extensions ORDER BY name;
EOF
🛠️ 高级配置和优化
内存优化配置
# 根据服务器内存调整配置
sudo nano /usr/local/pgsql/data/postgresql.conf
# 内存相关优化(8GB内存服务器示例)
shared_buffers = 2GB # 内存的25%
work_mem = 16MB # 每个查询的工作内存
maintenance_work_mem = 512MB # 维护操作内存
effective_cache_size = 6GB # 可用于缓存的磁盘空间估计值
# 性能优化
max_connections = 200 # 根据应用需求调整
checkpoint_completion_target = 0.9
random_page_cost = 1.1 # SSD设为1.1,HDD设为4.0
effective_io_concurrency = 200 # SSD可以设高,HDD设低
# WAL优化
wal_buffers = 16MB
min_wal_size = 1GB
max_wal_size = 4GB
配置归档和备份
# 创建归档目录
sudo mkdir -p /usr/local/pgsql/archive
sudo chown postgres:postgres /usr/local/pgsql/archive
# 配置归档
sudo nano /usr/local/pgsql/data/postgresql.conf
# 启用归档
archive_mode = on
archive_command = 'test ! -f /usr/local/pgsql/archive/%f && cp %p /usr/local/pgsql/archive/%f'
archive_timeout = 3600 # 每小时强制归档
📊 监控和维护脚本
创建监控脚本
# 创建数据库健康检查脚本
sudo tee /usr/local/bin/check_postgres.sh << 'EOF'
#!/bin/bash
# PostgreSQL健康检查脚本
PGHOME=/usr/local/pgsql
PGDATA=/usr/local/pgsql/data
PGPORT=5432
echo "=== PostgreSQL健康检查 $(date) ==="
# 检查服务状态
systemctl is-active postgresql > /dev/null 2>&1
if [ $? -eq 0 ]; then
echo "✓ 服务状态: 运行中"
else
echo "✗ 服务状态: 停止"
exit 1
fi
# 检查连接
$PGHOME/bin/pg_isready -p $PGPORT -h localhost -U postgres
if [ $? -eq 0 ]; then
echo "✓ 数据库连接: 正常"
else
echo "✗ 数据库连接: 失败"
fi
# 检查磁盘空间
df -h $PGDATA | tail -1
# 检查数据库大小
echo "数据库大小:"
$PGHOME/bin/psql -h localhost -U postgres -d postgres -c "
SELECT
datname as \"数据库\",
pg_size_pretty(pg_database_size(datname)) as \"大小\"
FROM pg_database
ORDER BY pg_database_size(datname) DESC;
"
echo "=== 检查完成 ==="
EOF
sudo chmod +x /usr/local/bin/check_postgres.sh
sudo chown postgres:postgres /usr/local/bin/check_postgres.sh
创建备份脚本
# 创建备份脚本
sudo tee /usr/local/bin/backup_postgres.sh << 'EOF'
#!/bin/bash
# PostgreSQL备份脚本
BACKUP_DIR="/usr/local/pgsql/backup"
DATE=$(date +%Y%m%d_%H%M%S)
RETENTION_DAYS=7
# 创建备份目录
mkdir -p $BACKUP_DIR/$DATE
echo "开始备份: $(date)"
# 备份所有数据库
/usr/local/pgsql/bin/pg_dumpall -h localhost -U postgres \
| gzip > $BACKUP_DIR/$DATE/full_backup_$DATE.sql.gz
# 备份单个重要数据库(可选)
/usr/local/pgsql/bin/pg_dump -h localhost -U postgres test_db \
-F c -f $BACKUP_DIR/$DATE/test_db_$DATE.dump
echo "备份完成: $(date)"
echo "备份文件:"
ls -lh $BACKUP_DIR/$DATE/
# 清理旧备份
find $BACKUP_DIR -type d -mtime +$RETENTION_DAYS -exec rm -rf {} \;
echo "已清理超过${RETENTION_DAYS}天的备份"
EOF
sudo chmod +x /usr/local/bin/backup_postgres.sh
sudo chown postgres:postgres /usr/local/bin/backup_postgres.sh
🎯 安装其他稳定版本
安装PostgreSQL 15.7(LTS版本)
cd /opt/postgresql_src
wget https://ftp.postgresql.org/pub/source/v15.7/postgresql-15.7.tar.gz
tar -zxvf postgresql-15.7.tar.gz
cd postgresql-15.7
mkdir build && cd build
# 配置(与16.2类似)
../configure \
--prefix=/usr/local/pgsql15 \
--with-pgport=5433 \
--with-openssl \
--with-perl \
--with-python
make -j$(nproc)
sudo make install
安装PostgreSQL 14.12(长期支持)
cd /opt/postgresql_src
wget https://ftp.postgresql.org/pub/source/v14.12/postgresql-14.12.tar.gz
tar -zxvf postgresql-14.12.tar.gz
cd postgresql-14.12
mkdir build && cd build
../configure --prefix=/usr/local/pgsql14 --with-openssl
make -j$(nproc)
sudo make install
⚠️ 故障排除
常见问题解决
# 1. 编译错误:缺少依赖
# 重新安装所有依赖后清理重试
make distclean
# 2. 启动失败:端口被占用
netstat -tlnp | grep 5432
sudo lsof -i :5432
# 3. 连接失败:认证问题
# 检查pg_hba.conf配置
# 4. 内存不足:调整编译并行度
make -j2 # 使用更少的并行任务
# 5. 查看详细错误日志
tail -f /usr/local/pgsql/data/log/postgresql-*.log
📝 总结
安装步骤回顾
- 安装依赖:确保编译环境完整
- 下载源码:选择稳定版本(推荐16.2或15.7)
- 配置编译:根据需求选择编译选项
- 编译安装:使用并行编译提高速度
- 初始化数据库:设置数据目录和编码
- 配置参数:优化性能和安全
- 创建服务:使用systemd管理
- 验证测试:确保安装成功
版本选择建议
- 生产环境:PostgreSQL 15.x(LTS支持到2027年)
- 学习测试:PostgreSQL 16.x(最新特性)
- 兼容性要求:PostgreSQL 14.x(支持到2026年)
关键配置文件
/usr/local/pgsql/data/postgresql.conf- 主配置文件/usr/local/pgsql/data/pg_hba.conf- 客户端认证/etc/systemd/system/postgresql.service- 服务文件
通过源码安装可以获得更好的性能和定制性,适合生产环境和对性能有要求的场景。
配置 PostgreSQL 主从复制(远程 → 本地 macOS 备份)
我将帮你配置 qiaopan.tech(远程服务器)作为主库,本地 macOS 作为从库进行异步复制,用于数据备份。
📋 架构设计
主库 (远程服务器) 从库 (本地 macOS)
qiaopan.tech:5432 → 本地:5432
↓
定时同步备份
🔧 第一部分:配置远程主库(qiaopan.tech)
步骤1:创建复制用户
# 在远程服务器上执行
sudo -i -u postgres psql
-- 创建复制专用用户
CREATE USER replicator WITH REPLICATION LOGIN
PASSWORD 'YourStrongReplicaPass123!'
CONNECTION LIMIT 3
VALID UNTIL '2025-12-31';
-- 验证创建
\du replicator
步骤2:配置主库 postgresql.conf
# 远程服务器上编辑配置文件
sudo nano /etc/postgresql/16/main/postgresql.conf
# 修改或添加以下配置
listen_addresses = '*' # 允许所有IP连接(或指定具体IP)
port = 5432 # 默认端口
wal_level = replica # 复制级别
max_wal_senders = 10 # 最大WAL发送进程数
wal_keep_size = 1024 # 保留1GB的WAL日志
max_replication_slots = 10 # 最大复制槽数
hot_standby = on # 从库可读
archive_mode = on # 开启归档
archive_command = 'test ! -f /var/lib/postgresql/16/main/archive/%f && cp %p /var/lib/postgresql/16/main/archive/%f'
# 性能优化
synchronous_commit = off # 异步复制,性能更好
wal_sender_timeout = 60s # 发送超时时间
步骤3:配置客户端认证(pg_hba.conf)
sudo nano /etc/postgresql/16/main/pg_hba.conf
# 添加以下行(根据你的本地IP配置)
# 允许本地macOS连接进行复制
# TYPE DATABASE USER ADDRESS METHOD
host replication replicator your_local_ip/32 md5
# 如果本地IP经常变化,可以放宽限制(不推荐生产环境)
# host replication replicator 0.0.0.0/0 md5
# 获取你的公网IP(在本地macOS执行)
curl ifconfig.me
步骤4:重启远程PostgreSQL并创建复制槽
# 重启服务
sudo systemctl restart postgresql@16-main
# 创建复制槽(可选,但推荐)
sudo -i -u postgres psql -c "SELECT * FROM pg_create_physical_replication_slot('macos_backup_slot');"
# 查看复制槽
sudo -i -u postgres psql -c "SELECT * FROM pg_replication_slots;"
# 查看当前WAL位置
sudo -i -u postgres psql -c "SELECT pg_current_wal_lsn();"
🔧 第二部分:配置本地macOS从库
步骤1:准备本地环境
# 停止本地PostgreSQL(如果正在运行)
/Library/PostgreSQL/16/bin/pg_ctl -D /Library/PostgreSQL/16/data stop
# 备份现有数据(如果重要)
sudo cp -r /Library/PostgreSQL/16/data /Library/PostgreSQL/16/data.backup.$(date +%Y%m%d)
# 删除现有数据目录
sudo rm -rf /Library/PostgreSQL/16/data
步骤2:从主库同步基础数据
# 使用pg_basebackup从远程主库同步数据
sudo -u postgres /Library/PostgreSQL/16/bin/pg_basebackup \
-h qiaopan.tech \
-p 5432 \
-U replicator \
-D /Library/PostgreSQL/16/data \
-Fp \ # 普通格式
-Xs \ # 使用流式复制
-R \ # 生成recovery配置
-P \ # 显示进度
-v # 详细输出
# 输入密码:YourStrongReplicaPass123!
步骤3:配置从库恢复参数
# 查看生成的配置
cat /Library/PostgreSQL/16/data/standby.signal
cat /Library/PostgreSQL/16/data/postgresql.auto.conf
# 如果未自动生成,手动创建
sudo tee /Library/PostgreSQL/16/data/standby.signal << EOF
standby_mode = 'on'
EOF
# 创建恢复配置
sudo tee /Library/PostgreSQL/16/data/postgresql.auto.conf << EOF
primary_conninfo = 'host=qiaopan.tech port=5432 user=replicator password=YourStrongReplicaPass123! application_name=macos_backup'
primary_slot_name = 'macos_backup_slot'
restore_command = 'cp /var/lib/postgresql/archive/%f %p'
recovery_target_timeline = 'latest'
hot_standby = 'on'
EOF
# 设置权限
sudo chown -R postgres:staff /Library/PostgreSQL/16/data
步骤4:配置本地postgresql.conf
sudo nano /Library/PostgreSQL/16/data/postgresql.conf
# 从库特有配置
hot_standby = on
hot_standby_feedback = on
max_standby_streaming_delay = 30s
max_standby_archive_delay = 30s
wal_receiver_status_interval = 10s
# 设置为只读模式(备份用途)
default_transaction_read_only = on
# 监听配置(本地使用)
listen_addresses = 'localhost'
port = 5432
步骤5:启动本地从库
# 启动PostgreSQL
sudo -u postgres /Library/PostgreSQL/16/bin/pg_ctl \
-D /Library/PostgreSQL/16/data \
-l /Library/PostgreSQL/16/logfile \
start
# 查看启动日志
tail -f /Library/PostgreSQL/16/logfile
🔍 验证复制状态
在远程主库上检查
# 查看复制连接状态
sudo -i -u postgres psql -c "
SELECT
pid,
usename,
application_name,
client_addr,
state,
sync_state,
pg_wal_lsn_diff(pg_current_wal_lsn(), sent_lsn) as sent_lag_bytes,
pg_wal_lsn_diff(pg_current_wal_lsn(), write_lsn) as write_lag_bytes,
pg_wal_lsn_diff(pg_current_wal_lsn(), flush_lsn) as flush_lag_bytes,
pg_wal_lsn_diff(pg_current_wal_lsn(), replay_lsn) as replay_lag_bytes
FROM pg_stat_replication;
"
# 查看复制槽状态
sudo -i -u postgres psql -c "SELECT * FROM pg_replication_slots;"
在本地从库上检查
# 检查恢复状态
/Library/PostgreSQL/16/bin/psql -U postgres -c "
SELECT
pg_is_in_recovery() as is_in_recovery,
pg_last_wal_receive_lsn() as last_receive_lsn,
pg_last_wal_replay_lsn() as last_replay_lsn,
pg_last_xact_replay_timestamp() as last_replay_time,
now() - pg_last_xact_replay_timestamp() as replay_delay,
pg_wal_lsn_diff(pg_last_wal_receive_lsn(), pg_last_wal_replay_lsn()) as replay_lag_bytes;
"
# 测试数据同步
# 在主库创建一个测试表
# ssh到远程服务器执行:
# sudo -i -u postgres psql -c "CREATE DATABASE backup_test;"
# sudo -i -u postgres psql -d backup_test -c "CREATE TABLE test_data AS SELECT generate_series(1,1000) as id, md5(random()::text) as data;"
# 在本地验证数据
/Library/PostgreSQL/16/bin/psql -U postgres -d backup_test -c "SELECT COUNT(*) FROM test_data;"
📊 创建监控和管理脚本
监控脚本:check_replication_status.sh
sudo tee /usr/local/bin/check_replication_status.sh << 'EOF'
#!/bin/bash
# 检查主从复制状态
PRIMARY_HOST="qiaopan.tech"
PRIMARY_PORT="5432"
PRIMARY_USER="postgres"
REPLICA_DATA="/Library/PostgreSQL/16/data"
echo "=== PostgreSQL 复制状态检查 $(date) ==="
echo "主库: ${PRIMARY_HOST}:${PRIMARY_PORT}"
echo "从库: localhost:5432"
# 检查从库状态
echo -e "\n1. 从库状态:"
if /Library/PostgreSQL/16/bin/pg_isready -h localhost; then
echo "✓ 从库服务正常"
# 检查是否在恢复模式
RECOVERY_STATUS=$(/Library/PostgreSQL/16/bin/psql -U postgres -t -c "SELECT pg_is_in_recovery();" 2>/dev/null)
if [ "$RECOVERY_STATUS" = "t" ]; then
echo "✓ 运行在恢复模式(从库)"
# 检查复制延迟
echo -e "\n2. 复制延迟:"
/Library/PostgreSQL/16/bin/psql -U postgres -c "
SELECT
now() - pg_last_xact_replay_timestamp() AS replication_delay,
pg_size_pretty(pg_wal_lsn_diff(pg_last_wal_receive_lsn(), pg_last_wal_replay_lsn())) AS replay_lag,
pg_last_xact_replay_timestamp() AS last_replay_time;
"
else
echo "✗ 不在恢复模式,可能不是从库"
fi
else
echo "✗ 从库服务异常"
fi
# 检查主库连接(可选)
echo -e "\n3. 主库连接测试:"
if nc -z $PRIMARY_HOST $PRIMARY_PORT 2>/dev/null; then
echo "✓ 可以连接到主库端口"
else
echo "✗ 无法连接到主库端口"
fi
echo -e "\n=== 检查完成 ==="
EOF
sudo chmod +x /usr/local/bin/check_replication_status.sh
定时备份验证脚本
sudo tee /usr/local/bin/verify_backup.sh << 'EOF'
#!/bin/bash
# 验证备份数据的完整性
LOG_FILE="/Library/PostgreSQL/16/logs/backup_verify.log"
REPLICA_DATA="/Library/PostgreSQL/16/data"
# 创建日志目录
mkdir -p /Library/PostgreSQL/16/logs
echo "$(date): 开始备份验证" | tee -a $LOG_FILE
# 1. 检查从库状态
if ! /Library/PostgreSQL/16/bin/pg_isready -h localhost; then
echo "$(date): 错误: 从库服务未运行" | tee -a $LOG_FILE
exit 1
fi
# 2. 检查恢复模式
RECOVERY_STATUS=$(/Library/PostgreSQL/16/bin/psql -U postgres -t -c "SELECT pg_is_in_recovery();" 2>/dev/null)
if [ "$RECOVERY_STATUS" != "t" ]; then
echo "$(date): 警告: 不在恢复模式" | tee -a $LOG_FILE
fi
# 3. 检查最近的复制活动
LAST_REPLAY=$(/Library/PostgreSQL/16/bin/psql -U postgres -t -c "SELECT pg_last_xact_replay_timestamp();" 2>/dev/null)
if [ -n "$LAST_REPLAY" ]; then
REPLAY_AGE=$(/Library/PostgreSQL/16/bin/psql -U postgres -t -c "SELECT EXTRACT(EPOCH FROM (now() - pg_last_xact_replay_timestamp()));" 2>/dev/null)
if [ $(echo "$REPLAY_AGE > 300" | bc) -eq 1 ]; then
echo "$(date): 警告: 最近5分钟内没有复制活动" | tee -a $LOG_FILE
else
echo "$(date): 正常: 最近有复制活动 ($REPLAY_AGE 秒前)" | tee -a $LOG_FILE
fi
fi
# 4. 检查关键数据库的完整性
echo "$(date): 检查数据库完整性" | tee -a $LOG_FILE
/Library/PostgreSQL/16/bin/psql -U postgres -t -c "
SELECT
datname,
pg_size_pretty(pg_database_size(datname)) as size,
pg_stat_file('base/' || oid || '/PG_VERSION') IS NOT NULL as files_ok
FROM pg_database
WHERE datname NOT IN ('template0', 'template1', 'postgres')
ORDER BY pg_database_size(datname) DESC;
" | tee -a $LOG_FILE
# 5. 检查表空间
echo -e "\n$(date): 表空间状态:" | tee -a $LOG_FILE
/Library/PostgreSQL/16/bin/psql -U postgres -t -c "
SELECT
spcname,
pg_tablespace_location(oid) as location,
pg_stat_file(pg_tablespace_location(oid)) IS NOT NULL as accessible
FROM pg_tablespace;
" | tee -a $LOG_FILE
echo "$(date): 备份验证完成" | tee -a $LOG_FILE
EOF
sudo chmod +x /usr/local/bin/verify_backup.sh
🔄 配置定时任务(crontab)
# 编辑当前用户的crontab
crontab -e
添加以下内容:
# PostgreSQL 备份监控和验证
# 每30分钟检查一次复制状态
*/30 * * * * /usr/local/bin/check_replication_status.sh >> /Library/PostgreSQL/16/logs/replication_monitor.log 2>&1
# 每天凌晨2点验证备份完整性
0 2 * * * /usr/local/bin/verify_backup.sh
# 每周一凌晨3点清理旧日志
0 3 * * 1 find /Library/PostgreSQL/16/logs -name "*.log" -mtime +30 -delete
# 每小时发送一次健康报告到通知(可选)
# 0 * * * * /usr/local/bin/send_health_report.sh
🛠️ 故障排除和恢复
问题1:网络中断后的恢复
# 如果网络中断导致复制停止
# 1. 检查主库是否可访问
ping -c 3 qiaopan.tech
# 2. 在从库重新启动复制
/Library/PostgreSQL/16/bin/pg_ctl -D /Library/PostgreSQL/16/data restart
# 3. 如果仍然失败,可能需要重新同步
/Library/PostgreSQL/16/bin/pg_ctl -D /Library/PostgreSQL/16/data stop
sudo rm -rf /Library/PostgreSQL/16/data
sudo -u postgres /Library/PostgreSQL/16/bin/pg_basebackup -h qiaopan.tech -p 5432 -U replicator -D /Library/PostgreSQL/16/data -Fp -Xs -R -P
/Library/PostgreSQL/16/bin/pg_ctl -D /Library/PostgreSQL/16/data start
问题2:复制延迟过大
# 查看当前延迟
/Library/PostgreSQL/16/bin/psql -U postgres -c "
SELECT
now() - pg_last_xact_replay_timestamp() as replication_delay,
pg_size_pretty(pg_wal_lsn_diff(pg_last_wal_receive_lsn(), pg_last_wal_replay_lsn())) as replay_lag;
"
# 如果延迟过大,可能是网络或性能问题
# 优化本地macOS PostgreSQL配置
sudo nano /Library/PostgreSQL/16/data/postgresql.conf
# 增加以下优化参数
max_standby_streaming_delay = -1 # 禁用延迟,尽快应用
max_standby_archive_delay = -1 # 禁用归档延迟
wal_receiver_timeout = 120s # 增加接收超时
问题3:主库空间不足
# 在主库检查WAL日志大小
sudo -i -u postgres psql -c "
SELECT
slot_name,
pg_size_pretty(pg_wal_lsn_diff(pg_current_wal_lsn(), restart_lsn)) as replication_lag,
active
FROM pg_replication_slots;
"
# 如果从库长时间离线,可能需要清理复制槽
# sudo -i -u postgres psql -c "SELECT pg_drop_replication_slot('macos_backup_slot');"
📱 创建简单的Web监控界面(可选)
# 安装Python依赖
pip3 install flask psycopg2-binary
# 创建监控应用
sudo tee /Library/PostgreSQL/16/scripts/monitor_app.py << 'EOF'
#!/usr/bin/env python3
from flask import Flask, jsonify
import psycopg2
import psycopg2.extras
import os
from datetime import datetime
app = Flask(__name__)
def get_replication_status():
"""获取复制状态"""
try:
conn = psycopg2.connect(
host="localhost",
port=5432,
user="postgres",
database="postgres"
)
cursor = conn.cursor(cursor_factory=psycopg2.extras.DictCursor)
# 检查是否在恢复模式
cursor.execute("SELECT pg_is_in_recovery() as is_replica")
is_replica = cursor.fetchone()['is_replica']
if is_replica:
# 获取复制延迟信息
cursor.execute("""
SELECT
now() - pg_last_xact_replay_timestamp() as replication_delay,
pg_wal_lsn_diff(pg_last_wal_receive_lsn(), pg_last_wal_replay_lsn()) as replay_lag_bytes,
pg_last_xact_replay_timestamp() as last_replay_time,
pg_last_wal_receive_lsn() as last_receive_lsn,
pg_last_wal_replay_lsn() as last_replay_lsn
""")
replication_info = dict(cursor.fetchone())
# 计算延迟秒数
if replication_info['replication_delay']:
replication_info['replication_delay_seconds'] = replication_info['replication_delay'].total_seconds()
else:
replication_info['replication_delay_seconds'] = None
replication_info['status'] = 'replicating'
else:
replication_info = {'status': 'not_replica'}
conn.close()
return replication_info
except Exception as e:
return {'status': 'error', 'message': str(e)}
@app.route('/api/replication/status')
def replication_status():
"""返回复制状态API"""
return jsonify(get_replication_status())
@app.route('/api/databases')
def list_databases():
"""列出所有数据库"""
try:
conn = psycopg2.connect(
host="localhost",
port=5432,
user="postgres",
database="postgres"
)
cursor = conn.cursor(cursor_factory=psycopg2.extras.DictCursor)
cursor.execute("""
SELECT
datname,
pg_size_pretty(pg_database_size(datname)) as size,
datcollate,
pg_stat_file('base/' || oid || '/PG_VERSION') IS NOT NULL as files_ok
FROM pg_database
WHERE datname NOT IN ('template0', 'template1')
ORDER BY pg_database_size(datname) DESC
""")
databases = []
for row in cursor:
databases.append(dict(row))
conn.close()
return jsonify({'databases': databases})
except Exception as e:
return jsonify({'error': str(e)})
@app.route('/')
def index():
"""简单监控页面"""
status = get_replication_status()
html = f"""
<html>
<head>
<title>PostgreSQL 备份监控</title>
<style>
body {{ font-family: Arial, sans-serif; margin: 40px; }}
.status {{ padding: 20px; border-radius: 5px; }}
.healthy {{ background-color: #d4edda; color: #155724; }}
.warning {{ background-color: #fff3cd; color: #856404; }}
.error {{ background-color: #f8d7da; color: #721c24; }}
</style>
</head>
<body>
<h1>PostgreSQL 备份监控</h1>
<p>时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}</p>
<div class="status {'healthy' if status.get('status') == 'replicating' else 'error'}">
<h2>复制状态: {status.get('status', 'unknown')}</h2>
"""
if status.get('status') == 'replicating':
html += f"""
<p>复制延迟: {status.get('replication_delay_seconds', 'N/A')} 秒</p>
<p>最后复制时间: {status.get('last_replay_time')}</p>
<p>WAL 接收位置: {status.get('last_receive_lsn')}</p>
<p>WAL 重放位置: {status.get('last_replay_lsn')}</p>
"""
elif status.get('status') == 'error':
html += f"<p>错误: {status.get('message')}</p>"
html += """
</div>
<p><a href="/api/replication/status">JSON 状态</a> |
<a href="/api/databases">数据库列表</a></p>
</body>
</html>
"""
return html
if __name__ == '__main__':
app.run(host='127.0.0.1', port=5000, debug=False)
EOF
# 创建启动脚本
sudo tee /Library/PostgreSQL/16/scripts/start_monitor.sh << 'EOF'
#!/bin/bash
cd /Library/PostgreSQL/16/scripts
nohup python3 monitor_app.py > monitor.log 2>&1 &
echo $! > monitor.pid
EOF
sudo chmod +x /Library/PostgreSQL/16/scripts/start_monitor.sh
📝 安全建议
使用SSH隧道(推荐):
# 在本地macOS创建SSH隧道 ssh -N -L 5433:localhost:5432 user@qiaopan.tech # 然后在配置中使用 localhost:5433 连接加强密码安全:
- 使用强密码
- 定期更换复制用户密码
- 使用不同密码用于不同用途
限制访问IP:
# 在主库pg_hba.conf中只允许你的公网IP host replication replicator your_public_ip/32 md5
✅ 完成验证
测试一切正常工作:
# 1. 检查复制状态
/usr/local/bin/check_replication_status.sh
# 2. 在主库创建测试数据
# ssh到远程服务器执行:
# sudo -i -u postgres psql -c "CREATE TABLE test_sync AS SELECT generate_series(1,100) as id, now() as timestamp;"
# 3. 在本地验证数据
/Library/PostgreSQL/16/bin/psql -U postgres -c "SELECT COUNT(*) FROM test_sync;"
# 4. 查看监控页面(如果启用了Web监控)
# 访问 http://localhost:5000
